Keywords: Bio-signals analysis and interpretation, Quantification of physiological parameters for diagnosis and treatment assessment
Abstract: In this talk, I review our recent studies on DNB (Dynamical Network Biomarkers) that provide early warning signals of imminent codim-1 local bifurcation from a healthy state to a disease state through a pre-disease state that is called Mibyo in Japanese and Weibing in Chinese. I also explain a possible application of DNB for early medicine by using real biological data such as gene expression data of mouse models as well as human. Further, I introduce recent collaborative work in our Moonshot project to apply control theory to possible early treatments of pre-disease states before transition to disease states.

Keywords: Intelligent transportation
Abstract: Effective timetable scheduling strategies are essential for passenger satisfaction in urban rail transit networks. Most existing passenger-centric timetable scheduling approaches generate a timetable according to deterministic passenger origin-destination (OD) demands. As passenger OD demands in urban rail transit networks generally show a high level of uncertainty, an effective timetable scheduling approach should take the uncertain passenger flows into account to generate a reliable timetable. In this paper, a scenario-based model predictive control (SMPC) approach is presented to handle uncertain passenger flows based on a passenger absorption model, where uncertainties are captured by several representative scenarios according to historical data. In each SMPC step, the optimization problem for generating the timetable can be formulated as a mixed-integer linear programming (MILP) problem, which can be solved efficiently by existing MILP solvers. A probabilistic performance level can be then determined based on the performance of SMPC under the representative scenarios. The effectiveness of the developed approach is evaluated through simulations on a part of the Beijing urban rail transit network.

Keywords: Intelligent transportation
Abstract: We consider the problem of a single Autonomous Vehicle (AV) merging into traffic consisting only of Human Driven Vehicles (HDVs) with the goal of minimizing both the travel time and energy consumption of the entire group of vehicles involved in the merging process. This is done by controlling only the AV and determining both the optimal merging sequence and the optimal AV trajectory associated with it. We derive an optimal index policy which prescribes the merging position of the AV within the group of HDVs. We also specify conditions under which the optimal index corresponds to the AV merging before all HDVs or after all HDVs, in which case no interaction of the AV with the HDVs is required. Simulation results are included to validate the optimal index policy and demonstrate cases where optimal merging can be achieved without requiring any explicit assumptions regarding human driving behavior.

Keywords: Intelligent transportation, Urban mobility, Connected vehicles
Abstract: In this paper, platoons of autonomous vehicles in urban road networks are considered. From a methodological point of view, the problem consists in characterizing vehicle state trajectory tubes by means of routing decisions complying with traffic congestion criteria. To this end, a novel distributed control architecture is conceived by taking advantage of two methodologies: the routing decisions are obtained by using a distributed reinforcement learning algorithm that exploits available traffic data and a bank of model predictive controllers is in charge of computing the more adequate control action for each involved vehicle.

Keywords: Urban mobility, Game theories, Control in economics
Abstract: This paper proposes a route optimization algorithm for a mixed package-passenger system that assigns the task of package delivery to taxicab drivers carrying passengers. The goal is to compensate for the shortage of manpower among package delivery drivers and to reduce CO2 emissions from delivery vehicles. Then, we propose a stable ridesharing matching scheme that achieves fare reductions for taxicab passengers. In addition, we formulate it as a mixed integer linear programming problem to determine the delivery task assignment to the driver and the optimal route, and propose a distributed method based on Lagrangian relaxation to reduce the computational effort. Finally, the effectiveness of the proposed algorithm is shown through numerical simulations.

Keywords: Urban mobility, Energy and distribution management systems, Smart parking
Abstract: Abstract: Over the last decade, many cities in the world have been facing the impact of urban population growth and rapid e-commerce spread on freight volumes and consequently on the number of road freight vehicles. These dynamics have fostered the central role of last-mile logistics. The transport sector is responsible for around 25% of total GHG global emissions, 30% of which are related to freight road transport. Urban freight transportation has remarkable implications in terms of air pollution, noise, and road security. In this context, the electrification of urban fleets could represent a viable and efficient solution to mitigate the environmental footprint of last-mile logistics. Furthermore, last-mile logistics also involves high organization costs and time inefficiencies for transportation firms and customers. The technological development of routing processes through a new optimized IT system (e.g., by means of digital twins) may play a key role in “greening” the last-mile logistic sector. In this research, we consider a case study of investments in Electric Vehicles, aiming at assessing their environmental and monetary costs and benefits, and the scalability of such a policy.

Keywords: Logistics; Traffic management; Fuel efficiency; Minimize travel time and CO2 consumption and price; Parking control.

Keywords:
Abstract: Shotgun presentations from Networked and Distributed Control (Interactive Session). Each talk will be for a 2 minutes duration.

Keywords:
Abstract: Shotgun presentations from Nonlinear Control Systems (Interactive Session). Each talk will be for a 2 minutes duration.

Keywords:
Abstract: Shotgun presentations from System Identification (Interactive Session). Each talk will be for a 2 minutes duration.

Keywords:
Abstract: Shotgun presentations from Data-driven and Numerical Optimal Control (Interactive Session). Each talk will be for a 2 minutes duration.

Keywords:
Abstract: Shotgun presentations from SAutonomous Vehicles (Interactive Session). Each talk will be for a 2 minutes duration.

Keywords:
Abstract: Shotgun presentations from Automotive Sensing and Control (Interactive Session). Each talk will be for a 2 minutes duration.

Keywords: Dynamic networks
Abstract: For consistent identification of a target module in a dynamic network with the local direct method, basically two prime conditions need to be satisfied: (a) a set of structural conditions on the choice of the predictor model, i.e. a set of input and output node variables, and (b) conditions on data-informativity. While for conditions (a) constructive algorithms for node selection have been presented that appropriately guarantee that the identified object can indeed reveal the target module, the requirements for satisfying (b) have not yet been integrated fully. In this paper, we will present simplified path-based results for generic data-informativity, and show how they can be integrated in constructive algorithms for predictor model selection that provide consistent target module estimates. It is shown that data-informativity not only requires a sufficient number of external excitation signals to be present in the network, but also puts restrictions on the structure of the predictor model, i.e. the selection of input and output node variables. Some examples are presented to illustrate the new results.

Keywords: Machine learning, Time series modelling, Continuous time system estimation
Abstract: Artificial intelligence (AI) models are increasingly ubiquitous in daily life, their unparalleled predictive and decision-making capabilities are being utilized for applications of all magnitudes, ranging from minor decisions to those with significant impacts on individuals and society. However, many of these models feature a multitude of redundant parameters, which render them incomprehensible to human understanding. The lack of transparency raises concerns about the reliability and fairness of the decisions made by AI models motivating a new field of research called eXplainable AI (XAI), which aims to elucidate complex AI model outcomes and develop tools to enable human understanding. Given the increasing impact of machine learning in the fields of data-driven estimation and control, it becomes crucial to integrate XAI tools with control theory to better comprehend the decisions made by AI models in estimation and control. In this article, we propose the use of an XAI method called Local Interpretable Model-Agnostic Explanations (LIME) to explain the mechanisms behind a black-box estimation algorithm processing time-series data. Moreover, we demonstrate that LIME can be used to identify a local linearized model that approximates the complex machine learning algorithm. We show that the identified local linearized model can shed light on the dynamic of the model that generated the training data.

Keywords: Nonlinear system identification, Dynamic networks, Identification for control
Abstract: The complex, multi-variable, highly nonlinear and strong coupling characteristics of water distribution systems (WDSs) has significantly limited the capability of model-based approaches for control purposes in such systems. With the emerging application of high-resolution metering devices and historical data, model-free identification of WDSs can facilitate the control design without tedious modeling complexities. This paper develops a data-driven framework to facilitate the identification of nonlinear models of WDSs using available data. A quadruple tank system that represents the nonlinear and strong coupling nature of WDSs is considered as the test system and sparse identification of nonlinear dynamics (SINDy) is utilized to identify the nonlinear dynamics from the data. Unlike existing modeling approaches that either heavily rely on knowing the detailed dynamics of the system (model-based) or designs that relay on large historical data and are not interpretable (data-driven approaches), the proposed model-free identification framework is parsimonious, which can accurately capture the dynamics of the quadruple tank process with available measurements suitable for control problems. The effectiveness of the proposed approach is validated using time-domain simulations in MATLAB.

Keywords: Identifiability, Dynamic networks, Experiment design
Abstract: This paper presents identifiability conditions for identifying the complete dynamics of diffusively coupled linear networks. These conditions are derived by exploiting the uniqueness of the nonmonic polynomial network description, given the locations of the actuators and sensors. The analysis is performed under a more relaxed instrumentation setup than the typical restriction to a full set of sensors (full measurement) or a full set of actuators (full excitation). This leads to more general identifiability conditions, including more flexible instrumentation requirements.

Keywords: Data-driven control, Learning for control, Control of networks
Abstract: When creating models of a system, there is always a tradeoff between the ease of modelling a part and the increased value it brings to the model. Learning a model using machine learning, we might have less control over what dynamics to capture, but can also capture things we don't necessarily understand. Using universal differential equations we can combine the two, taking scientific models and embedding machine learning into them, with the goal of giving us the best of both worlds. In this paper, we extend an existing model with small neural networks to capture missing dynamics. The specific use-case involves a microservice fluid model, where the learned extension improves the range of parameters the combined model can reliably produce predictions over. This means the model can be reused in a wider parameter space before needing to be retrained. We also explore the possibility of imposing bias on the network based on features of the model, and how that affects performance.

Keywords: Dynamic networks, Closed loop identification, Identifiability
Abstract: In a dynamic network of interconnected transfer functions, it is not necessary to use all the node signals for estimating a local transfer function. Given the network topology, detailed conditions are available for selecting inputs and outputs in a (MIMO) predictor model that warrants consistent and minimum variance estimation of a target module through the so-called local direct method. Motivated by the existing minimum-input signal selection approach that gradually incorporates additional signals, an alternative graphical algorithm for signal selection is developed in this work by directly exploiting the complete network graph. Then, as a straightforward application of existing analytical results, graphical conditions for consistent identification are derived for the novel signal selection approach. We show by an example that in some cases, for the consistent estimation of the target module, the developed method leads to fewer selected signals than the original minimum-input method.

Keywords: Adaptive control, Stochastic optimal control problems, Time-delay systems
Abstract: In this paper, we present a time-delay approach to extremum seeking (ES) corrupted by white noise for uncertain static quadratic maps. We first apply a time-delay approach to the ES system and arrive at a neutral type time-delay system with stochastic perturbations. Then we further present the latter system as a retarded one and employ the variation of constants formula for the mean-square exponential ultimate boundedness analysis. Under the assumption that the upper bound of the 6th moment of the estimation error is a known arbitrarily large constant L, explicit conditions in terms of simple scalar inequalities depending on the bound L, tuning parameters and the intensity of measurement noise are established to guarantee the mean-square exponential ultimate boundedness of the ES control systems. Comparatively to the existing results for ES with measurement noise via the qualitative analysis, our approach can provide a quantitative analysis and simplify the stability analysis process as well. A numerical simulation is given to illustrate the efficiency of the proposed method.

Keywords: Data-driven robust control, Robust learning systems, Networked systems
Abstract: Formulating state estimation for a large-scale, discrete-time, linear time-invariant system as a least-squares problem can be computationally challenging as the problem dimensions increase with time. Recently, randomized sampling has demonstrated promising results in approximating this problem by using fewer rows, resulting in a polynomial-sized approximate problem. However, these algorithms necessitate calculating the statistical leverage scores of the rows, which can be challenging. In this paper, we propose an alternative approach to approximate leverage scores by uniformly sampling rows. We demonstrate that this method provides a sufficiently weak form of approximation for obtaining an estimate of each leverage score. Our approach delivers a reasonable approximation of the leverage scores, which is suitable for approximating and solving the least-squares estimation problem with proven theoretical guarantees.

Keywords: Robustness analysis, Time-invariant systems, Stability of nonlinear systems
Abstract: In this paper, we focus on the lower bounds of the L_{p} (pin[1,infty), p=infty) induced norms of continuous-time LTI systems where input signals are restricted to be nonnegative. This induced norm, called the L_{p+} induced norm, is particularly useful for the stability analysis of nonlinear feedback systems constructed from linear systems and static nonlinearities where the nonlinearities provide only nonnegative signals for the case p=2. To have deeper understanding on the L_{p+} induced norm, we analyze its lower bounds with respect to the standard L_{p} induced norm in this paper. As the main result, we show that the L_{p+} induced norm of an LTI system cannot be smaller than the L_{p} induced norm scaled by 2^{(1-p)/p} for in [1,infty) (scaled by 2^{-1} for p=infty). On the other hand, in the case where p=2, we further propose a method to compute better (larger) lower bounds for single-input systems via reduction of the lower bound analysis problem into a semi-infinite programming problem. The effectiveness of the lower bound computation method, together with an upper bound computation method proposed in our preceding paper, is illustrated by numerical examples.

Keywords: Control of hybrid systems, Data-based control, Quantized control
Abstract: This paper is concerned with the controller synthesis problem for discrete-time unknown systems against safety specifications via control barrier certificates. Typically, control barrier certificates provide sufficient conditions for the satisfaction of safety specifications by separating the safe and unsafe regions of the system. By synthesizing these certificates in conjunction with control policies, one is able to keep the system safe. In our work, we parameterize the control barrier certificates and corresponding control policies as neural networks and learn them simultaneously by utilizing finitely many data samples obtained from the unknown system. We derive a so-called validity condition to formally verify the obtained certificates and integrate this condition within the training framework to achieve provably correct guarantees at the end of training time. In particular, we exploit Lipschitz continuity properties of the neural networks and utilize robust training techniques to ensure that the trained networks not only satisfy the required control barrier certificate conditions across the finitely many training data samples but over the entire state set. We then demonstrate the effectiveness of our approach with the help of a case study.

Keywords: Output regulation, Data-driven robust control, Controller constraints and structure
Abstract: While reinforcement learning has made great improvements, state-of-the-art algorithms can still struggle with seemingly simple set-point feedback control problems. One reason for this is that the learned controller may not be able to excite the system dynamics well enough initially, and therefore it can take a long time to get data that is informative enough to learn for good control. The paper contributes by augmentation of reinforcement learning with a simple guiding feedback controller, for example, a proportional controller. The key advantage in set point control is a much improved excitation that improves the convergence properties of the reinforcement learning controller significantly. This can be very important in real-world control where quick and accurate convergence is needed. The proposed method is evaluated with simulation and on a real-world double tank process with promising results.

Keywords: Data-driven robust control, Convex optimization, Linear systems
Abstract: This paper applies a polynomial optimization-based framework towards the superstabilizing control of an Autoregressive with Exogenous Input (ARX) model given noisy data observations. The recorded input and output values are corrupted with L-infinity-bounded noise where the bounds are known. This is an instance of Error in Variables (EIV) in which true internal state of the ARX system remains unknown. The consistency set of ARX models compatible with noisy data has a bilinearity between unknown plant parameters and unknown noise terms. The requirement for a dynamic compensator to superstabilize all consistent plants is expressed using polynomial nonnegativity constraints, and solved using sum-of-squares (SOS) methods in a converging hierarchy of semidefinite programs in increasing size. The computational complexity of this method may be reduced by applying a Theorem of Alternatives to eliminate the noise terms. The effectiveness of this method is demonstrated on control of example ARX models.

Keywords: Controller constraints and structure, Sliding mode control, Singularities in optimization
Abstract: In recent years, safety-critical control has attracted much attention because it provides safety for the target system and a new perspective for designing a controller according to a control objective. This paper considers a safety-critical control technique to obtain a simple procedure for redesigning discontinuous control. First, we revisit an optimal control problem for a quadratic-formed performance function without input costs and propose a procedure for remaking a discontinuous optimal controller into a continuous quasi-optimal controller via the technique of safety-critical control. Then, we also try deriving a simple control design procedure for a chained system via the concept of safety-critical control. Finally, the validity is confirmed by numerical simulation.

Keywords: Sliding mode control, Aerospace, Stability of nonlinear systems
Abstract: Since the dynamic model of spacecraft has the characteristics of non-linear, kinematic couplings, uncertainties and nonstationary disturbance, it has become a challenging problem to accurately control the relative position and attitude of the spacecraft. A radial basis function neural network (RBFNN)-based sliding mode controller (SMC) is proposed for trajectory tracking of spacecraft autonomous proximity in this paper. Firstly, a six degree-of-freedom (DOF) relative motion dynamics model is developed for close proximity operations. The modified Rodrigues parameters are applied to solve the problem of singularity. Then, a SMC that does not require accurate model information is designed. RBFNN is used to adaptively eliminated the model uncertainty impacts on the system. Finally, the stability of the relative motion dynamics is proved via Lyapunov stability theory. Simulation results illustrate that the method can attenuate the attitude and position errors, reduce the chattering of the input and decrease the overshoot of the control torque effectively.

Keywords: Sliding mode control, Nonlinear system identification, Data-driven control
Abstract: The Koopman operator framework allows for a linear, but infinite-dimensional, representation of the dynamics of a non-linear system. The Koopman modes, or observables, and the resulting linear dynamics are derived purely using a data-driven framework, where the data are system outputs measured at discrete samples; improving accuracy of the Koopman representation requires a large number of such modes to be considered. Recent results consider the system input as well, in the derivation of the discrete linear dynamics, thus enabling the design of controllers. Sliding mode controllers (SMCs), including the discrete-time versions, can handle parameter uncertainties and variations and also ensure that the control objective is satisfied in finite time. In this paper, a discrete-time SMC is designed for the output control of a dynamic system approximated by fewer Koopman modes; the SMC is expected to handle uncertainties introduced by the ignored modes. Conditions are identified for the closed-loop system to be stable, with the occurrence of sliding mode.

Keywords: Sliding mode control, Adaptive control, Sampled-data control
Abstract: This paper proposes a model-free sampled-data sliding mode control of piezoelectric actuators considering time delay. As time delay is a challenging subject in controller design, we try to eradicate its effect on the tracking performance of the control system as much as possible. For this goal, an equivalent sampled-data data-based model from the continuous-time nonlinear system of a piezoelectric actuator is first derived. Then, we include the delay of the control system into the sliding surface to enrich the controller performance. Next, the sliding mode controller is designed using the equivalent model and the delayed sliding surface. No observer is used to compensate for the total disturbances. Thus, the scheme is easy to design and implement. The stability of the closed-loop system is theoretically studied, and the controller performance is confirmed through comparative experimental tests. Based on the experimental results, the proposed controller successfully obtains control objectives.

Keywords: Sliding mode control, Switching stability and control, Convex optimization
Abstract: This paper presents a time-discretization methodology for designing a control law for stabilizing a continuous plant. It relies on the concept of immersion and invariance principle with implicit-Euler discretization of Twisting algorithm. The multi-valuedness of the signum function on the off-the-manifold is fully utilized by this discretization. By eliminating numerical chattering in the control input and system output, the resulting controller ensures a smooth stabilization of the off-the-manifold. Additionally, the exogenous disturbance that is currently affecting the system is diminished by a factor of h (h is the sampling interval) as well. The efficacy of proposed technique are shown through comparative study in the simulations.

Keywords: Sliding mode control, Stability of nonlinear systems, Nonlinear observers and filter design
Abstract: Accurate and robust positioning of rotor axle is essential for efficient and safe operation of high-speed rotational machines with active magnetic bearings. This study presents a cascaded nonlinear control strategy for vertical axial positioning of an active magnetic bearing system. The proposed scheme employs two sliding mode controllers for regulating rotor vertical position and current and an adaptive estimator to invert the nonlinear input mapping. Uniform asymptotic stability is proven for the closed-loop system and the efficacy and performance of the proposed design is evaluated in simulation.

Keywords: Fault-tolerant, Diagnosis, UAVs
Abstract: This paper addresses the active fault tolerant control problem for a tilted hexarotor under both actuator faults and failures, modelled respectively as partial or total losses of effectiveness. A disturbance observer based control approach is adopted in order to estimate faults and failures, and cope with them at the control level. An inner-outer loop tracking control law is then designed which, thanks to the actuation properties of the hexarotor, does not require any reconfiguration neither in response to a fault nor failure. The control scheme is tested in simulation, with parameters of a real hexarotor and realistic sensor noise.

Keywords: Nonlinear observers and filter design, UAVs, Information and sensor fusion
Abstract: Inertial Navigation Systems (INS) are algorithms that fuse inertial measurements of angular velocity and specific acceleration with supplementary sensors including GNSS and magnetometers to estimate the position, velocity and attitude, or extended pose, of a vehicle. The industry-standard extended Kalman filter (EKF) does not come with strong stability or robustness guarantees and can be subject to catastrophic failure. This paper exploits a Lie group symmetry of the INS dynamics to propose the first nonlinear observer for INS with error dynamics that are almost-globally asymptotically and locally exponentially stable, independently of the chosen gains. The observer is aided only by a GNSS measurement of position. As expected, the convergence guarantee depends on persistence of excitation of the vehicle's specific acceleration in the inertial frame. Simulation results demonstrate the observer's performance and its ability to converge from extreme errors in the initial state estimates.

Keywords: UAVs, Aerospace
Abstract: This work focuses on transition trajectory generation for a special class of hybrid vertical takeoff and landing (VTOL) vehicles called quadrotor biplane tailsitter. The transition refers to the change of flight modes for a hybrid VTOL vehicle. A transition maneuver is challenging as the vehicle undergoes in low-speed high angle of attack condition which is highly unstable. This work proposes methods to generate an optimal trajectory for the forward transition of the vehicle. A simplified 3-degree-of-freedom (3-DOF) model of the vehicle is considered for generating the transition trajectory. The optimization problem is solved using nonlinear programming with appropriate dynamic and physical constraints. The work considers multiple optimality conditions that include minimal control effort, minimal transition time, and a combination of them. The results of different approaches are compared considering the feasibility of implementation on a physical vehicle.

Keywords: UAVs, Event-triggered and self-triggered control, Robustness analysis
Abstract: This paper proposes a self-triggered control policy for a general kind of nonlinear system, ensuring robust performance under various perturbations and alleviating communication burden simultaneously. A novel fully actuated multirotors applied this method shows great robustness to perturbations and is able of tracking six dimensional decoupled trajectories without continuously background sensor monitoring. Rigorous input-to-state stability analysis is deduced and numerical simulations conrm the validity of the policy.

Keywords: UAVs, Flying robots, Navigation, guidance and control
Abstract: The usage of a drone in pursuit has been researched quite extensively, but with only one drone, the pursuit suffers from sensory limitations where sensors are either expensive or have limited capabilities. In this paper, a drone swarm controller system for pursuit is developed to solve the problem based on Parallel Navigation (PN) and Proportional-Integral-Derivative (PID). The controller consists of two main parts with different functions, each in charge of navigation and control modulation. Navigation is performed by a High-Level Controller built on PN, which allows navigation toward an object while avoiding obstacles simultaneously. Control modulation is performed by a Low-Level Controller based on a PID controller. The implementation study is then performed in a simulated environment with varying swarm members. Results show that the controller system successfully coordinated the drone swarm to surround the target after it is found, and pursuit with more drones consistently locates the target faster.

Keywords: UAVs, Disturbance rejection, Stability of nonlinear systems
Abstract: In this paper, we apply finite time settling control to a hierarchical linearized quadcopter model and verify its robustness against model errors by simulation. Hierarchical linearization is one of the exact linearization method and is linearizable on all points. Finite time settling control has the property of finite-time stability, which perfectly matches the equilibrium point in a finite time range. This property increases robustness near the equilibrium point. Thus, it is expected to improve the stability of drones that are vulnerable to wind disturbances and model errors. As a comparison, linear state feedback control with exponential stability is used. The feedback gain is calculated using linear quadratic regulator(LQR), which is known to have high robustness. The simulation results show that the finite time settling controller has higher the trajectory tracking performance than the linear quadratic regulator.

Keywords: Autonomous mobility, Scheduling and optimization of transportation systems, Intelligent Transportation Systems
Abstract: The advent of vehicle autonomy, connectivity and electric powertrains is expected to enable the deployment of Autonomous Mobility-on-Demand systems. Crucially, the routing and charging activities of these fleets are impacted by the design of the individual vehicles and the surrounding charging infrastructure which, in turn, should be designed to account for the intended fleet operation. This paper presents a modeling and optimization framework where we optimize the activities of the fleet jointly with the placement of the charging infrastructure. We adopt a mesoscopic planning perspective and devise a time-invariant model of the fleet activities in terms of routes and charging patterns, explicitly capturing the state of charge of the vehicles by resampling the road network as a digraph with iso-energy arcs. Then, we cast the problem as a mixed-integer linear program that guarantees global optimality and can be solved in less than 10 min. Finally, we showcase two case studies with real-world taxi data in Manhattan, NYC: The first one captures the optimal trade-off between charging infrastructure prevalence and the empty-mileage driven by the fleet. We observe that jointly optimizing the infrastructure siting significantly outperforms heuristic placement policies, and that increasing the number of stations is beneficial only up to a certain point. The second case focuses on vehicle design and shows that deploying vehicles equipped with a smaller battery results in the lowest energy consumption: Although necessitating more trips to the charging stations, such fleets require about 12% less energy than the vehicles with a larger battery capacity.

Keywords: Electric and solar vehicles, Nonlinear and optimal automotive control, Design methodologies
Abstract: This paper introduces a framework to systematically optimize the control and design of an electric vehicle transmission, connecting powertrain sizing approaches with low-level gearbox design methods. To this end, we first create analytical models of the individual transmission components: gears, shafts, bearings, clutches, and synchronizers. Second, we construct a transmission by systematically configuring a topology with these components. Third, we integrate the composed transmission within a powertrain and vehicle model, and compute the minimum-energy control strategies and design, employing two solution algorithms, namely dynamic programming and exhaustive search, respectively, providing global optimality guarantees. Finally, we carry out the control and design optimization of a one- and two-speed transmission for a small electric family car, whereby we observe that a two-speed transmission can improve the energy consumption by 0.5% with respect to a one-speed transmission, whilst also satisfying the gradeability and performance requirements.

Keywords: Electric and solar vehicles, Energy and distribution management systems, Smart grids
Abstract: The increase in the number of Electric Vehicles (EVs), results in additional stress on the grid system to fulfill the EVs charging demand. However, with EVs integration into the wholesale electricity market, EVs fleet operators can play a big role to balance the grid by providing ancillary services and bidding in the electricity markets. Therefore, optimal strategies for bidding in the wholesale energy markets using EVs are becoming a necessity. In this paper, a bi-level optimization model was developed for scheduling the charging of EVs as well as allocating optimal bid capacity in the day-ahead, intraday, and reserve balancing markets to maximize the profit of EV fleet operators from the charging and discharging of the EVs. The results demonstrate that the model is successful in charging EVs and obtaining remarkable revenue streams from participating in multiple electricity markets.

Keywords: Electric and solar vehicles, Extremum seeking and model free adaptive control, Safety and security in transportation systems
Abstract: To ensure braking stability and maximize energy recovery, electric vehicles combine two braking modes, friction and regenerative braking. This paper proposes a novel braking torque distribution (BTD) algorithm between frictional/regenerative brake systems. Safety and regeneration energy are the two major issues in the design of a braking torque allocation. Based on the extremum-seeking (ES) technique, the multiple objectives and constraints issues of the hybrid ABS braking torque distribution are well addressed. The offered BTD can maximize the regeneration energy and improve the battery’s state of charge (SOC) by optimizing the electric braking system efficiency, while respecting the braking actuators constraints. The validity and effectiveness of the proposed BTD has been demonstrated by simulation in the MATLAB/Simulink environment.

Keywords: Cooperative logistics
Abstract: We introduce an aided decision framework for co-opetition, i.e., collaboration of competitor companies, in food distribution by considering the economic and environmental advantages of it. We show the benefits and issues related to co-opetition of the retailing companies in food distribution networks. The mathematical model indicates major financial and environmental improvements through increasing the filing rate of the trucks and reducing the number of trucks used in the logistics network as a result of co-opetition of the companies. The results are based on a case study of a food distribution network serving three retailers. This paper provides practical facts which might be useful for the logistics managers of the food retailing companies. The sustainable environmental benefits of co-opetition through the lens of cooperative game theory are investigated.

Keywords: Control architectures in automotive control, Hybrid and alternative drive vehicles, Human mechatronics
Abstract: Combining the advantages of standard bicycles and electrified vehicles, electric bikes (e-Bikes) are promising vehicles to reduce emission and traffic. The current literature on e-Bikes ranges from works on the energy management to the vehicle control to properly govern the human-vehicle interaction. This last point is fundamental in chain-less series bikes, where the link between the human and the vehicle behavior is only given by a control law. In this work, we address this problem in a series-parallel bike. In particular, we provide an extension of the virtual-chain concept, born for series bikes, and then we improve it developing a virtual-bike framework. Experimental results are used to validate the effectiveness of the solutions, when the cyclist is actually riding the bike.

Keywords: Adaptive control
Abstract: In this work, we consider the problem of regret minimization in adaptive minimum variance and linear quadratic control problems. Regret minimization has been extensively studied in the literature for both types of adaptive control problems. Most of these works give results of the optimal rate of the regret in the asymptotic regime. In the minimum variance case, the optimal asymptotic rate for the regret is log(T ) which can be reached without any additional external excitation. On the contrary, for most adaptive linear quadratic problems, it is necessary to add an external excitation in order to get the optimal asymptotic rate of √T . In this paper, we will actually show from a theoretical study, as well as, in simulations that when the control horizon is pre-specified a lower regret can be obtained with either no external excitation or a new exploration type termed immediate.

Keywords: Real-time optimal control, Constrained control
Abstract: In this work, we propose a control scheme for linear systems subject to pointwise in time state and input constraints that aims to minimize time-varying and a priori unknown cost functions. The proposed controller is based on online convex optimization and a reference governor. In particular, we apply online gradient descent to track the time-varying and a priori unknown optimal steady state of the system. Moreover, we use a lambda-contractive set to enforce constraint satisfaction and a sufficient convergence rate of the closed-loop system to the optimal steady state. We prove that the proposed scheme is recursively feasible, ensures that the state and input constraints are satisfied at all times, and achieves a dynamic regret that is linearly bounded by the variation of the cost functions. The algorithm's performance and constraint satisfaction is illustrated by means of a simulation example.

Keywords: Predictive control, Optimal control theory, Constrained control
Abstract: This paper presents a synthesis method for the generalised dynamic regret problem, comparing the performance of a strictly causal controller to the optimal non-causal controller under a weighted disturbance. This framework encompasses both the dynamic regret problem, considering the difference of the incurred costs, as well as the competitive ratio, which considers their ratio, and which have both been proposed as inherently adaptive alternatives to classical control methods. Furthermore, we extend the synthesis to the case of pointwise-in-time bounds on the disturbance and show that the optimal solution is no worse than the bounded energy optimal solution and is lower bounded by a constant factor, which is only dependent on the disturbance weight. The proposed optimisation-based synthesis allows considering systems subject to state and input constraints. Finally, we provide a numerical example which compares the synthesised controller performance to mathcal{H}_2- and mathcal{H}_infty-controllers.

Keywords: Control problems under conflict and/or uncertainties, Data-based control, Linear systems
Abstract: The setting of an agent making decisions under uncertainty and under dynamic constraints is common for the fields of optimal control, reinforcement learning, and recently also for online learning. In the online learning setting, the quality of an agent's decision is often quantified by the concept of regret, comparing the performance of the chosen decisions to the best possible ones in hindsight. While regret is a useful performance measure, when dynamical systems are concerned, it is important to also assess the stability of the closed-loop system for a chosen policy. In this work, we show that for linear state feedback policies and linear systems subject to adversarial disturbances, linear regret implies asymptotic stability in both time-varying and time-invariant settings. Conversely, we also show that bounded input bounded state stability and summability of the state transition matrices imply linear regret.

Keywords: Optimal control theory, Predictive control, Convex optimization
Abstract: We consider control of dynamical systems through the lens of competitive analysis. Most prior work in this area focuses on minimizing regret, that is, the loss relative to an ideal clairvoyant policy that has noncausal access to past, present, and future disturbances. Motivated by the observation that the optimal cost only provides coarse information about the ideal closed-loop behavior, we instead propose directly minimizing the tracking error relative to the optimal trajectories in hindsight, i.e., imitating the clairvoyant policy. By embracing a system level perspective, we present an efficient optimization-based approach for computing follow-the-clairvoyant (FTC) safe controllers. We prove that these attain minimal regret if no constraints are imposed on the noncausal benchmark. In addition, we present numerical experiments to show that our policy retains the hallmark of competitive algorithms of interpolating between classical H2 and H-infinity control laws – while consistently outperforming regret minimization methods in constrained scenarios thanks to the superior ability to chase the clairvoyant.

Keywords: Observer design, Parameter and state estimation, Robust estimation
Abstract: Kalman and H-infinity filters, the most popular paradigms for linear state estimation, are designed for very specific specific noise and disturbance patterns, which may not appear in practice. State observers based on the minimization of regret measures are a promising alternative, as they aim to adapt to recognizable patterns in the estimation error. In this paper, we show that the regret minimization problem for finite horizon estimation can be cast into a simple convex optimization problem. For this purpose, we first rewrite linear time-varying system dynamics using a novel system level synthesis parametrization for state estimation, capable of handling both disturbance and measurement noise. We then provide a tractable formulation for the minimization of regret based on semi-definite programming. Both contributions make the minimal regret observer design easily implementable in practice. Finally, numerical experiments show that the computed observer can significantly outperform both Kalman and H-infinity filters.

Keywords: Multi-agent systems, Decentralized control, Machine learning
Abstract: Multi-agent reinforcement learning (MARL) under partial observability is notoriously challenging as the agents only have asymmetric partial observations of the system. In this paper, we study MARL in decentralized partially observable Markov decision processes (Dec-POMDPs) with partial history sharing. In search of decentralized and tractable MARL solutions, we identify the appropriate conditions under which we can adopt the common information approach to naturally extend existing single-agent policy learners to Dec-POMDPs. In particular, under the conditions of bounded local memories and an efficient representation of the common information, we present a MARL algorithm that learns a near-optimal finite-memory policy in Dec-POMDPs. We establish the iteration complexity of the algorithm, which depends only linearly on the number of agents. Simulations on classic Dec-POMDP tasks show that our approach significantly outperforms existing decentralized solutions, and nearly matches the centralized ones that require stronger informational assumptions.

Keywords: Multi-agent systems, Coordination of multiple vehicle systems
Abstract: In this paper, we study the cooperative target-fencing problem for an n-dimensional target with a general trajectory. The goal is to fence the given target into the convex hull formed by a group of n-dimensional vehicles which are modeled by double-integrator systems. We propose a control law to accomplish this task while avoiding collisions among the vehicles and preserving the connectivity of the communication network. In particular, our approach can handle a target whose position trajectory is any twice continuously differentiable function. Moreover, our result does not rely on the velocity measurements of the vehicles. Thus, it is more practical compared with the existing approaches in the literature. The effectiveness of the proposed method is illustrated by a numerical example.

Keywords: Multi-agent systems, Linear systems, Optimal control theory
Abstract: In this paper, we investigate the distributed H2 suboptimal leader-follower consensus control problem for linear multi-agent systems using dynamic output feedback. By considering an autonomous leader, a number of followers, and an associated H2 cost functional, we aim to design a distributed protocol to ensure that the leader-follower consensus is achieved while the associated H2 cost is smaller than an a priori given upper bound. To this end, we first show that the H2 suboptimal leader-follower consensus control problem can be equivalently derived as the H2 suboptimal control problem of a set of independent systems. Based on this, we then present a design method for computing a distributed protocol. The computation of the feedback gains involves two Riccati inequalities whose dimension matches the state dimension of the agents. A simulation example is provided to demonstrate the performance of the proposed protocol.

Keywords: Multi-agent systems, Decentralized control and large-scale systems, Complex system management
Abstract: The broadcast control (BC) framework for multi-agent coordination tasks is suitable for a limited-sized multi-agent system, where a single coordinator broadcasts the same scalar signal to the performing agents without discrimination, and the agents do not communicate among themselves. The BC technique uses a stochastic optimization mechanism to control a group of agents for accomplishing a global shared task. However, as the number of agents increases, this single coordinator-based BC either requires a longer time to converge or fails, keeping its application scopes limited. This paper proposes an enhancement scheme of BC incorporating distributed multiple coordinators with their parting scopes to control larger multi-agent systems effectively. Specifically, considering a coverage control task of agents, a hierarchical distributed coordination scheme is introduced through Pseudo-perturbation based BC, where multiple coordinators broadcast the same feedback signal to the agents within their range. The proposed scheme is evaluated using numerical simulation, and results are compared with the typical BC scheme.

Keywords: Multi-agent systems, Control under communication constraints, Coordination of multiple vehicle systems
Abstract: This work discusses formation control of heterogeneous Multi-Agent Systems (MASs) under discrete time setting where its formation size is scalable by a scaling factor. The case with and without input constraints are discussed. The communication network is assumed to be jointly connected and the leader-follower network is assumed for the unconstrained case. Compared to the previous work on distributed scalable formation control for heterogeneous agents, this work considers input constraints in the formulation. The proposed algorithm is based on the discrete time version of the internal model principle and the constraint is handled by the command governor. Numerical examples are shown to illustrate the proposed method.

Keywords: Multi-agent systems, Control over networks, Distributed control and estimation
Abstract: Biased measurements in an inter-networked systems can have severe repercussions in closed-loop stability of the individual systems and decelerate dynamical consensus among the interacting agents. Bias in the measurement, even constant, cannot be dealt with ad hoc techniques of robust control, in the presence of additive perturbations, because the control gain amplifies the disturbance. One way to account for the effect of measurement bias is then to rely on adaptive control. This has been done in the literature in the context of individual systems, but to the best of our knowledge not for multi-agent systems, while ensuring consensus control. In this paper we provide a model-reference-adaptive-control scheme to ensure dynamic consensus of generic (stabilizable) linear systems interconnected over directed graphs and under the influence of constant bias measurements. Our controller ensures global asymptotic stability of the synchronization manifold and convergence of the bias estimates.

Keywords: Process optimisation, Constraint and security monitoring and control, Control of renewable energy resources
Abstract: The resonance condition for maximum absorption of the energy carried by ocean waves incident to an oscillating wave energy converter (WEC) is one of the most well known results of the WEC control research. It was shown that the maximum occurs when the WEC velocity is in phase with the wave excitation force. The condition was obtained with the assumption that the WEC frees to oscillate with whatever amplitude was necessary, and that the power take-off (PTO) system can deliver the unlimited force. In practical WEC implementations, this assumption does not hold. The purpose of this paper is twofold. First, to obtain the frequency response of the closed-loop WEC motion, and of the optimal control law in terms of the hydrodynamic coefficients with limited PTO force and limited motion amplitudes. Secondly, the obtained results are used to investigate the resonance condition of the WEC velocity with the incoming wave in the presence of constraints.

Keywords: Process optimisation, Maintenance scheduling and production planning, Applications in advanced materials manufacturing
Abstract: This work presents the sensitivity analysis of a hierarchical scheduling framework for the copper smelting process with respect to the change in matte grade. The aim is to find the matte grade that can maximize the throughput of the copper smelting process. This study considers multiple-PSC units and a multiple-batch hierarchical scheduling framework that finds an optimal schedule for the process using promising heuristics. The process is briefly discussed, and the simulation results are discussed in detail to provide a comprehensive view of the effects of the matte grade on the operation of the copper smelting process.

Keywords: Process optimisation, Data visualization
Abstract: The work is devoted to solving one of the important problems of design of gas field development - the problem of optimal placement of producing wells. We consider a method for solving the problem of dimensionality, which arose when the model became more complex due to the addition of reservoir filtration properties. The results of applying the method are presented.

Keywords: Process performance monitoring/statistical process control, Statistical methods/signal analysis for FDI, AI methods for FDI
Abstract: The theory and practice of process monitoring are diverging. Practical process monitoring requires fault detection, identification, diagnosis, and the implementation of process recovery actions. Algorithmic design and automation of all these steps are required if the future promise of more autonomous plants is to be realised. In contrast, theoretical research in data-driven process monitoring is overwhelmingly focused on fault detection. This paper discusses the current challenges of data-driven process monitoring research and presents a conceptual framework for improved experimentation of end-to-end process monitoring approaches. An end-to-end process monitoring solution is defined as the complete set of automated algorithms that is able to execute (in real-time) fault detection, fault identification, fault diagnosis, as well as process recovery intervention advisories. The major contribution of this framework is in the increased relevance added to the process monitoring problem to be solved, particularly in the extent of autonomy that is required by any proposed process monitoring solution.

Keywords: Real time optimization and control, Infinite-dimensional systems (linear case)
Abstract: Without repeated online optimization, optimal operation of uncertain distributed parameter systems (DPSs) is addressed through the proposed self-optimizing control (SOC) method, where the existing global SOC is expanded to a class of first-order systems. The main characteristic is that the SOC controlled variables (CVs) are designed offline with a limited number of sensors and actuators based on an infinite-dimensional system itself. The effectiveness of the proposed methods is demonstrated through a steam-jacketed heat transfer case.

Keywords: Real time optimization and control, Advanced control technology, Industrial applications of process control
Abstract: This paper presents bench-scale experiment results of a new coagulant dosage control for water purification processes. The developed new control scheme has a cascade structure where an inner loop PI controller works to determine coagulant dosing rate by using a flocculation image sensor and an outer loop set-point optimizer calculates the set-point of the inner loop controller using extremum seeking (ES) technique. The image sensor measures the electrophoretic velocity (EV) which is an index of flocculation process performance and its set-point is optimized by using a performance index consisting of a total operational cost and treated water quality constraint. Several bench-scale experiment results show that the developed controller works well for the case that the treated water quality constraint is active as well as for the case that the optimal point is an interior point of the feasible region.

Keywords: Advanced control technology, Industrial applications of process control, Control system design
Abstract: This paper describes the development and application of the Linear Quadratic Regulator (LQR) to the Selective Catalytic Reduction (SCR) process in Combined Cycle (CC) Power Plant. The whole closed-loop control is implemented by the actual plant Distributed Control System (DCS). To improve control dynamics, we are proposing cascade override controller system. The effectiveness of the designed controller is evaluated by not only the simulation study but also the actual plant operation data during commercial operation. The controller covers all the operation conditions, and has been continuously used from the commissioning stage to this day.

Keywords: Advanced control technology, Process control applications
Abstract: Bidirectional inventory control has been shown to solve the problem of maximizing the production of units arranged in series, through automatic reconfiguring of the inventory control loops, when temporary or permanent bottlenecks occur in any section of the process. This control system deals with constraints related to maximum flow, but minimum flow constraints are also typical in process systems to avoid improper operation. This work proposes an extension to the bidirectional inventory control structure that incorporates minimum flow constraints, through the use of additional level controllers with intermediate setpoints, and additional selector blocks. The order in which the selectors are implemented indicates the priority for giving up on the controlled variables, and the intermediate setpoint values affect how long the process can run in feasible operation. The proposed control structure successfully prevents constraint violation when the problem is feasible, retaining the reconfiguring properties of bidirectional inventory control.

Keywords: Process optimisation, Advanced process control
Abstract: Free-form bending is a bending technique, whereas the bending tool head combines both linear and rotary motions. One problem with free-form bending is that an incorrectly bent tube section can no longer be corrected. Therefore, if a defect is detected during bending, a strategy for bending the remaining tube sections must be determined in such a way that the overall shape of the tube is as close as possible to the desired one. In this paper, a solution based on optimal recalculation of the residual trajectories of the bending tool head is presented and dierent cost functions are discussed in terms of their eect on the overall shape. Simulation results have shown a signicant improvement of the resulting end shape of the workpiece upon subjecting it to perturbation.

Keywords: Advanced process control, Process optimisation, Model predictive and optimization-based control
Abstract: A model predictive controller was implemented at the flash drying unit of the Anglo American Platinum Polokwane smelter. The controller uses a mix of standard model predictive control technology and dynamic control limits (based on the rate of change of the hot gas generator average bed temperature) to improve temperature stability across the unit. The improved temperature control resulted in the ability to process 1.80% more concentrate through the unit without the need to increase coal feed, as well as reducing the number of coal feeder trips that result from high temperature events.

Keywords: Advanced process control, Nonlinear adaptive control, Data-based control
Abstract: Modern metal forming poses increasing requirements on the surface properties of metallic strips, as these determine not only appearance but also processability during subsequent processing steps such as deep drawing or painting. In this context, an online scheme for pass scheduling of a rolling mill is presented that optimizes the gap adjustments of both roll stands on a tandem rolling mill in order to independently control the strip’s geometric and surface properties, more precisely its thickness and mean surface roughness. The optimization integrates three process models describing the entire rolling process. All models are adapted online to address process disturbances, such as material property variations or the wear of the roughened work rolls. Model adaption is carried out based on sensor data using Gaussian process regression for model identification. Hereby, the control loop is closed by the online adaptation of the underlying process models. Finally, the adaptive pass scheduling is validated on a tandem rolling mill with a DC04 steel of dimensions 1mm x 8.1mm. The experimental results indicate high tracking performance of the proposed optimization scheme as the outgoing strip thickness is successfully kept constant at a level of 0.92mm +- 0.008mm for a height reduction of 8% while controlling the surface roughness between 1.1µm and 2.5µm.

Keywords: Advanced process control, Process control applications
Abstract: Control systems for secondary grinding circuits follow widely accepted guidelines in the mineral processing industry. However, some common practices appear to be built on questionable assumptions, and become detrimental to the variability of the product size distribution. Secondary grinding constitutes an intermediate unit operation reducing the diameter of ore particles to meet the requirements for downstream separation. Hence the importance of the product size distribution, which represents the only product attribute measured online directly related to actual economic performance of the overall plant. This paper demonstrates with a simulation case study how controlling the hydrocylone inlet pressure at a constant setpoint does not serve the sought purpose. It compares this base case scenario with alternative ones, aiming for a constant particle size index at the overflow. Results show that under hardness disturbances, a constant hydrocyclone inlet pressure doesnt lead to any benefit, and even induce additional variability to the product size index resulting from the adjustments made to the hydrocyclone feed flow rate.

Keywords: Model predictive and optimization-based control, Process control applications, Process modeling and identification
Abstract: Hydrogen-fueled vehicles are a promising technology for reducing CO2 emissions in the mobility sector. This paper presents a stochastic receding horizon controller for operating a real-world hydrogen refueling station in a cost-efficient manner, while considering uncertainty in the hydrogen demand. We derive grey-box models for the system and we use them to formulate a predictive control taking the form of a mixed-integer linear program. We solve it by relying on the scenario approach and we ensure a reduced computational footprint by resorting on a non-uniform prediction horizon discretization and on a Lagrangian-based decomposition method.

Keywords: Model predictive and optimization-based control, Process control applications
Abstract: Dual fluidized bed steam gasification enables the production of gaseous energy carriers from woody biomass or biogenic residues. The circulation of bed material in dual fluidized bed gasifiers strongly affects the process behavior. Therefore, precise control of the bed material circulation is desired. This paper presents a control algorithm addressing two aspects of the given problem setting: On the one hand, redundant control actuators are available. Typically, there are several air streams to the reactors influencing the bed material circulation. On the other hand, only black box models with uncertainties in their model parameters are available for model-based control design. The presented control algorithm uses a model predictive controller considering known uncertainties in the model parameters and drives the process in a region with the lowest model uncertainties. This results in an improvement of the closed-loop performance when the actual plant deviates from the internal model used for the model predictive control predictions. Simulations show 66 % less offset from the design trajectory with the presented algorithm when compared to a standard model predictive controller.

Keywords: Model predictive and optimization-based control, Process control applications, Process modeling and identification
Abstract: Rigorous process simulation has become a tool that academic and industrial environments are exploiting, mainly to extract information useful for maximizing profit. As a matter of fact, the detailed thermodynamic models contained in commercial or open-source software are able to represent the behavior of a chemical process far better than a linearized model. On the other hand, designing customized model predictive controllers (MPC) has proven to enhance process performance over traditional control architectures. Therefore, in this paper, we present the interaction of an easy-to-use MPC algorithm developed in Python with the rigorous simulator UniSim Design®. The communication exploits the UniSim Design® spreadsheets as the variables database to be read/written by Python, by stopping or not the simulation before every control action. The software communication has been properly developed so to maintain the flexibility of the original MPC code and to exploit different controller designs. Two different test cases are presented to show the effectiveness of the proposed methodology: a simple two-phase separator and a more complex debutanizer column. System identification is used to build the controller’s linear models, various MPC designs differing in considering disturbances as measurable have been analyzed and satisfactory results are obtained.

Keywords: Control of distributed systems, Model predictive and optimization-based control, Machine learning and data analytics in process control
Abstract: Many industrial processes consist of a series of interconnected components. While the components are often individually controllable, the potentially complex interaction mechanisms limit the design options for a process-wide controller. Typical reasons are that parts of the component interactions are difficult to model or that the overall, detailed physical model is very complex. Due to these challenges, using model-based control approaches, such as model predictive control, often becomes infeasible. To this end, an adaptive two-degree of freedom control concept for reference tracking tasks is proposed. A data-driven surrogate model of the coupled system is trained, focusing on the reference tracking objective, and exploiting the model structure. This surrogate model is used in a two-degree of freedom control concept, splitting the tasks into reference tracking and disturbance rejection. The proposed control concept applies to a broad class of reference tracking processes subject to partially unknown system dynamics. It is tested exemplarily to control an interconnected three-tank system.

Keywords: Application of power electronics, Control system design, Instrumentation and control systems
Abstract: This manuscript presents a novel direct current model predictive control with long prediction horizon for medium-voltage polyphase permanent magnet synchronous motor drives. By utilizing the proposed technique, we can reduce the armature current harmonics and torque ripples, minimize the electrical and mechanical losses, improve the torque and speed rendering quality, and achieve a highly reliable permanent magnet motor drive. The state-space model of a representative polyphase permanent magnet synchronous motor is derived with the armature current and flux linkage jointly chosen as the state-space variables. Leveraging completing the squares, the mixed-integer-quadratic-programming (MIQP) problem can be formulated and efficiently solved, which provides the optimal switching sequence for the IGBT inverter. The proposed control algorithm is also implemented and examined with computer simulation to show the robustness and effectiveness.

Keywords: Model predictive and optimization-based control, Machine learning and data analytics in process control, Advanced control technology
Abstract: There is a growing interest in developing an efficient data-driven control method that can be implemented into digitized manufacturing processes. Model-free reinforcement learning (RL) is a machine learning method that can directly learn the optimal control policy from the process data. However, the model-free RL shows higher cost variance than the model-based method and may require an infeasible amount of data to learn the optimal control policy. Motivated by the fact that the system identification to linear model shows high data efficiency and stable performance, this paper proposes combining the linear model predictive control (MPC) with Q-learning. This combined scheme, Q-MPC, can improve the control performance more stably and safely. For the case study, linear MPC, Q-MPC, DDPG, TD3, and SAC methods are applied to the nonlinear benchmark system, mainly focusing on the learning speed and cost variance.

Keywords: Control of renewable energy resources, Model predictive and optimization-based control, Constraint and security monitoring and control
Abstract: In the context of material synthesis and thermal stress tests, solar furnaces play a fundamental role. These experimental systems are usually operated manually by trained operators. However, the use of automatic controllers can be essential to take into account the different dynamics of the treated materials, follow the different temperature profiles required in the tests, meet system security requirements, and face the intermittency of the main energy source, the solar energy. On these bases, the present work proposes a control system for temperature regulation in solar furnaces. The controller is based on a nonlinear model predictive control strategy that deals, in a single control layer, with the nonlinear behaviour of the system, the effect of disturbances, and the system's operating constraints. The controller was experimentally tested in the solar furnace SF60 of Plataforma Solar de Almería (southeast Spain), showing adequate performance against setpoint temperature changes, reaching a settling time of the order of 54 % faster than the open-loop system's time constant, and effectively rejecting irradiance disturbances. In addition, the operational constraints were ensured in almost all operating circumstances, observing only slight deviations of less than 2 ºC. These results allow us to position the proposed controller as a relevant tool for daily operation.

Keywords: Control of renewable energy resources
Abstract: The thermoelectric generator is an essential device used in the process of thermoelectric power generation. This paper researches the energy optimization problem of thermoelectric generators, and the goal is to maximize total output energy over a duration of time. The Multi-Step Load Optimization (MSLO) method is developed for addressing the problem. In our proposed method, we use Artificial Bee Colony (ABC) algorithm to search for the resistance values of multi-step to find a high-quality resistance sequence swiftly. Next, the encoding scheme and local search operators are developed. Instead of using the simulation model directly, we use the Kriging model as evaluation function of ABC algorithm to speed up the calculation of the evaluation value. Results show that the kriging model has high efficiency and accuracy, and when compared to state-of-art maximum power point tracking (MPPT) algorithms, the MSLO method has better optimization performance.

Keywords: Control of renewable energy resources, Control system design, Application of power electronics
Abstract: Even if the electrical grid is subject to constrained disturbances, doubly-fed induction generator (DFIG)-based wind turbines should be able to synchronize their stator voltage with that of the grid to ensure a smooth connection to the electric power system. In order to face the ynchronization task under simultaneously unbalanced and harmonically distorted grid voltages, a complex-valued sliding-mode control (SMC) algorithm, naturally chatter-free and phase-locked loop (PLL)-independent, is proposed. By accomplishing a stationary reference frame-based design, decomposition into positive- and negative-sequences and harmonic components is not required. The finite-time convergence of such algorithm is analytically demonstrated when subject to both parametric and unmodeled uncertainties, as well as disturbances. Simulation over a 2-MW DFIG model has been carried out in order to validate the performance and robustness of the suggested control structure under unbalanced and harmonically distorted grid voltage, variable speed wind profile, substantial parameter deviations and grid frequency variation.

Keywords: Control of renewable energy resources, Optimal operation and control of power systems, Real time simulation and dispatching
Abstract: Due to the global trend towards net zero emissions and increasing environmental awareness, some renewable energy sources (RES) and electric vehicles (EVs) have been integrated into the power system. An energy management system (EMS) can effectively integrate and coordinate distributed energy resource (DER) control in response to grid changes and impacts. Proper regulation of renewable energy and electric vehicles can reduce microgrid operating costs, which are critical for improving the safety and stability of power systems. Therefore, this paper proposes a three-stage optimal scheduling strategy from the perspective of a microgrid operator (MGO) to efficiently utilize resources. Based on an electric energy storage system (ESS), the traditional centralized optimal control method is compared with the method proposed to calculate the time and operating cost. It is verified that the method proposed can indeed execute an ancillary service (AS). In addition, the model predictive control (MPC) employed by the ESS can significantly improve the robustness of the system and reduce the occurrence of exceeding contract capacity.

Keywords: Control of renewable energy resources, Control problems under conflict and/or uncertainties, Energy systems
Abstract: In this paper, an approach for active power control of individual wind turbines is presented. State-of-the-art controllers typically employ separate control loops for torque and pitch control. In contrast, we use a multivariable control approach. In detail, active power control is achieved by using reference trajectories for generator speed, generator torque, and pitch angle such that a desired power demand is met if weather conditions allow. Then, a linear quadratic (LQ) optimal controller is used for reference tracking. In an OpenFAST simulation environment, the controller is compared to a state-of-the-art approach. The simulations show a similar active power tracking performance, while the LQ optimal controller results in lower mechanical wear. Moreover, the presented approach exhibits good reference tracking and by improving the reference trajectory generation further performance increases can be expected.

Keywords: Control of renewable energy resources, Distributed fault-tolerant Control, Fault accommodation and Reconfiguration strategies
Abstract: Wind farms (WFs) are becoming more popular around the world, especially in offshore areas, and they require reliable and available wind turbines (WTs) in order to generate cost-effective wind power. To achieve this, there is a need for effective solutions to address faults and improve the reliability and availability of wind turbines and the overall wind farm. This paper proposes a fault-tolerant cooperative control (FTCC) method for wind farms, which uses a model predictive control (MPC) approach at the turbine level, supplemented by a control reallocation mechanism (CRM) at the farm level. This method can handle both mild and severe power-loss faults in turbines, as demonstrated by various simulation studies on an advanced wind farm benchmark. The proposed solutions are highly efficient and effective.

Keywords: Learning and adaptation in autonomous vehicles, Control problems under conflict and/or uncertainties, Data-based control
Abstract: Autonomous car racing is a challenging task, as it requires precise applications of control while the vehicle is operating at cornering speeds. Traditional autonomous pipelines require accurate pre-mapping, localization, and planning which make the task computationally expensive and environment-dependent. Recent works propose use of imitation and reinforcement learning to train end-to-end deep neural networks and have shown promising results for high-speed racing. However, the end-to-end models may be dangerous to be deployed on real systems, as the neural networks are treated as black-box models devoid of any provable safety guarantees. In this work we propose a decoupled approach where an optimal end-to-end controller and a state prediction end-to-end model are learned together, and the predicted state of the vehicle is used to formulate a control barrier function for safeguarding the vehicle to stay within lane boundaries. We validate our algorithm both on a high-fidelity Carla driving simulator and a 1/10-scale RC car on a real track. The evaluation results suggest that using an explicit safety controller helps to learn the task safely with fewer iterations and makes it possible to safely navigate the vehicle on the track along the more challenging racing line.

Keywords: Autonomous vehicles, Learning and adaptation in autonomous vehicles, Trajectory and path planning
Abstract: In recent years, imitation learning (IL) has been widely used in industry as the core of autonomous vehicle (AV) planning modules. However, previous IL works show sample inefficiency and low generalisation in safety-critical scenarios, on which they are rarely tested. As a result, IL planners can reach a performance plateau where adding more training data ceases to improve the learnt policy. First, our work presents an IL model using the spline coefficient parameterisation and offline expert queries to enhance safety and training efficiency. Then, we expose the weakness of the learnt IL policy by synthetically generating critical scenarios through optimisation of parameters of the driver's risk field (DRF), a parametric human driving behaviour model implemented in a multi-agent traffic simulator based on the Lyft Prediction Dataset. To continuously improve the learnt policy, we retrain the IL model with augmented data. Thanks to the expressivity and interpretability of the DRF, the desired driving behaviours can be encoded and aggregated to the original training data. Our work constitutes a full development cycle that can efficiently and continuously improve the learnt IL policies in closed-loop. Finally, we show that our IL planner developed with less training resource still has superior performance compared to the previous state-of-the-art.

Keywords: Learning and adaptation in autonomous vehicles, Sensing, Networks of robots and intelligent sensors
Abstract: Autonomous search using teams of multiple agents need tractable coordination strategies between the search agents. The strategy must lower the time to identify interesting areas in the search environment, lower the costs/energy usage by the search agents during movement and sensing, and be resilient to the noise present in the sensed data due to the use of low-cost and low-weight sensors. We propose a data-driven, multi-agent search algorithm to achieve these goals using the framework of thresholding multi-armed bandits. For our algorithm, we also provide finite upper bounds on the time taken to complete the search, on the time taken to label all interesting cells, and on the economic costs incurred during the search.

Keywords: Multi-vehicle systems, Learning and adaptation in autonomous vehicles, Cooperative navigation
Abstract: Accurate and efficient speed prediction is crucial in autonomous driving or intelligent driving assistance systems to improve vehicle trajectory prediction accuracy, optimal decision planning, and energy management strategies. However, the traffic scene constructed by ego-vehicle and surrounding vehicles is a dynamic process, its complex spatial-temporal characteristics increase the complexity and challenge of predicting vehicle speed. This paper proposes a new method to solve the problem of using spatial-temporal characteristics in the process of vehicle speed prediction. The graph structure is used to describe the vehicle interaction scene and reflect the spatial relationship between vehicles. A temporal dynamic graph convolutional network (TDGCN) is proposed to predict vehicle speed by using temporal and spatial characteristics. The network combines graph convolutional neural network (GCN) and long short-term memory neural network (LSTM). GCN can process the complex topological structure in graph data to extract the spatial characteristics of traffic scenes, and LSTM can process the temporal characteristics of dynamic changes in traffic scenes. Finally, the TDGCN model is used to predict the vehicle speed based on the real driving dataset NGSIM I-80. The simulation results show that when predicting the speed after 1 frame and 10 frames, the root means a square error of the prediction results is 2.051 and 2.086, and the acceleration and deceleration changes of the vehicle can be correctly reflected, which proves the effectiveness of the TDGCN model in predicting the vehicle speed by using the spatial-temporal characteristics.

Keywords: Autonomous vehicles, Learning and adaptation in autonomous vehicles
Abstract: This paper introduces a learning-based Nonlinear Model Predictive Control (NMPC) method that combines NMPC with a Reinforcement Learning (RL) algorithm to achieve automatic parameter tuning of the NMPC optimizer, resulting in better control performance. In this study, two learning-based models were designed based on the tabular Q-learning algorithm but with different definitions of state and action spaces. To test the effectiveness of the proposed model, we conducted two kinds of experiments in which the models were applied to optimize the lane-keeping performance of an autonomous driving system. The case study results from simulations showed that the agent could match a proper parameter matrix for the NMPC within one minute. In real-world experiments, we extended the proposed control scheme to practical driving tasks using a 1:8 scale Audi model car in a specific experimental field. The model exhibited acceptable robustness in the face of relatively large deviations from the sensors and other real-time interference. These results demonstrate that the proposed learning-based NMPC method is a promising direction for solving real-time control problems.

Keywords: Intelligent Transportation Systems, Multi-agent systems, Consensus and reinforcement learning control
Abstract: Mobility systems often suffer from a high price of anarchy due to the uncontrolled behavior of selfish users. This may result in societal costs that are significantly higher compared to what could be achieved by a centralized system-optimal controller. Monetary tolling schemes can effectively align the behavior of selfish users with the system-optimum. Yet, they inevitably discriminate the population in terms of income. Artificial currencies were recently presented as an effective alternative that can achieve the same performance, whilst guaranteeing fairness among the population. However, those studies were based on behavioral models that may differ from practical implementations. This paper presents a data-driven approach to automatically adapt artificial-currency tolls within repetitive-game settings. We first consider a parallel-arc setting whereby users commute on a daily basis from an individual origin to an individual destination, choosing a route in exchange of an artificial-currency price or reward, while accounting for the impact of the choices of the other users on travel discomfort. Second, we devise a model-based reinforcement learning controller that autonomously learns the optimal pricing policy by interacting with the proposed framework considering the closeness of the observed aggregate flows to a desired system-optimal distribution as a reward function. Our numerical results show that the proposed data-driven pricing scheme can effectively align the users' flows with the system optimum, significantly reducing the societal costs with respect to the uncontrolled flows (by about 15% and 25% depending on the scenario), and respond to environmental changes in a robust and efficient manner.

Keywords: Game theories
Abstract: The allocation of heterogeneous resources plays an increasingly important role in the completion of system objectives and tasks. In this paper, we focus on deriving optimal strategies for the allocation of heterogeneous resources in a well-known adversarial model known as the General Lotto game. In standard formulations, outcomes are determined solely by the players' allocation strategies of a single type of resource across multiple contests. In particular, a player wins a contest if it sends more resources than the opponent. Here, we propose a multi-resource extension where the winner of a contest is now determined not only by the amount of resources allocated, but also by the composition of resource types that are allocated. We completely characterize the equilibrium payoffs and strategies for two distinct formulations. The first consists of a weakest-link/best-shot winning rule, and the second considers a winning rule based on a linear combination of the allocated resources. We then provide equilibrium investment strategies in scenarios where the resource types are costly to purchase, and players each have a limited monetary budget.

Keywords: Game theories
Abstract: The General Lotto game is a well-studied model where two opposing players strategically allocate their resources to multiple battlefields. In its classic setting, each player's objective is to secure as much value as possible by winning many individual battlefields. However, in many applications, a player must secure particular subsets of battlefields in order to succeed. The classic setting fails to capture a variety of adversarial interactions -- for instance, ensuring the security of networks or cyber-security domains. In this paper, we focus on a particular alternate player objective known as the majoritarian objective, where a player needs to secure a majority of battlefields in order to succeed. Equilibrium characterizations for the majoritarian objective in the existing literature are limited to particular symmetric settings where both players place the same value in succeeding. Our contributions extend these equilibrium solutions to asymmetric cases in two different settings: 1) both players have fixed and asymmetric resource budgets, and 2) both players place different values in succeeding and pay costs for allocating resources.

Keywords: Game theories, Optimal control theory, Climate change impact and adaptation measures
Abstract: One of the most widely used models for studying the geographical economics of climate change is the Regional Integrated model of Climate and the Economy (RICE). In this paper, we investigate how cooperation and non-cooperation arise in climate policy across regions under the RICE framework from the standpoints of game theory and optimal control. We show that the RICE model is inherently a dynamic game. We study both cooperative and non-cooperative solutions to this RICE dynamic game. Our results show how game theory may be used to help international negotiations reach an agreement on regional climate-change mitigation strategies, as well as how cooperative and competitive regional relationships impact future climate change.

Keywords: Game theories, Differential or dynamic games, Switching stability and control
Abstract: It is common that each agent in a noncooperative system has multiple objectives but the stability property for Nash equilibria in such a game has seldomly been studied. To solve this, we illustrate the characterization of Nash equilibria in the noncooperative systems with quadratic vector payoff functions analytically. It turns out that the Nash equilibria of such systems can be characterized by a set. Depending on the parameters of the payoff functions, we investigate the property of the Nash equilibrium set and present some sufficient conditions where the set is compact and connected. Furthermore, we consider the pseudo-gradient dynamics for the agents and present a sufficient condition where the Nash equilibrium set is asymptotically stable. Several numerical examples are presented to illustrate our results.

Keywords: Game theories, Networks (structural properties), Control problems under conflict and/or uncertainties
Abstract: This paper investigates a game-theoretic model of agents interacting on a network to compare the efficiency of direct transfers to participants and changing the interaction network characteristics. We study the effect of various control policies - individual and structural interventions - in network games with linear best-response. It is shown that control of network characteristics can be more effective than homogeneous targeting intervention.

Keywords: Industrial applications of process control, Process modeling and identification, Machine learning and data analytics in process control
Abstract: Oil sand is one of the main oil resources in Canada. In oil sands operation, primary separation cell, where bitumen is separated from the sand using water-based gravity separation, generates froth, middling and tailing. Tailing contains solids, water, and residual bitumen. The presence of bitumen in the tailing leads to loss of bitumen and increases environmental footprints. In order to remove residual bitumen, tailing is sent to a flotation cell. Determining the bitumen content in the froth of the flotation cell is essential for ensuring the efficiency of the process operation and improving environmental performance. Thus, samples of the flotation froth are often collected over a relatively long period of time and then sent to a laboratory for analysis. While laboratory analysis can provide a more accurate measurement of cumulative bitumen content over the sample collection period, it is costly and time-consuming. The long processing time in the laboratory will hinder real-time applications. In recent years, computer vision technology has been widely used to monitor and control flotation processes. This article proposes a computer vision-based solution for estimating cumulative bitumen content in flotation froth that usually contains a relatively small amount of bitumen. In spite of the frequent recording of images of the froth, the corresponding bitumen content label (laboratory analysis result) is not commonly available. The froth is collected batch by batch, and there is only one laboratory analysis of bitumen content in each batch, which reflects the cumulative bitumen content collected over the entire batch. Furthermore, images of the froth are often degraded by noise and varying lighting conditions. To solve the problem, a Kalman filter-based algorithm is proposed to restore the contaminated images. Gray-level co-occurrence matrix (GLCM) is then used to extract color and textural features from the restored images. Using these features, a linear regression model is built for predicting cumulative bitumen content over each batch. Validation results in a laboratory-scale flotation process demonstrate the proposed algorithm is promising in estimating bitumen content.

Keywords: Machine learning and data analytics in process control, Process modeling and identification, Model predictive and optimization-based control
Abstract: This work develops a Lyapunov-based economic model predictive control (LEMPC) scheme that utilizes recurrent neural networks (RNNs) with online update to optimize the economic benefits of switched nonlinear systems subject to a prescribed switching schedule. We first develop an initial offline-learning RNN using historical operational data, and then update RNN models using real-time data to improve model prediction accuracy. The generalized error bound for RNNs updated online with non-independent and identically distributed (non-i.i.d.) data samples is first derived. Subsequently, by incorporating the online update of RNNs within LEMPC, probabilistic closed-loop stability and economic optimality are achieved simultaneously for switched nonlinear systems accounting for the RNN generalized error bound. A chemical process example with scheduled mode transitions is used to demonstrate that the closed-loop economic performance under LEMPC can be improved using online learning of RNNs.

Keywords: Machine learning and data analytics in process control, Machine learning methods and applications, Process control applications
Abstract: Membrane-filtration processes play a crucial role in modern pharmacy and medicine but also in food industry and in biotechnology. During filtration, the filtered particles increasingly block the filter membrane leading to a reduced throughput up to a total blockage over time. Forecasting this membrane fouling would allow appropriate countermeasures to be taken in time to slow down the blockage and thus increase the overall yield. This paper introduces the challenges of yield maximization in membrane-filtration processes despite membrane fouling and possible advantages in forecasting membrane fouling. Further, the paper discusses the difficulties in the membrane-filtration process data and its pre-processing which need to be addressed to realize forecasting membrane fouling using data-driven machine learning and data analytics in process control.

Keywords: Machine learning and data analytics in process control, Model predictive and optimization-based control, Process control applications
Abstract: This work presents a framework for developing physics-informed recurrent neural network (PIRNN) models and PIRNN-based predictive control schemes for batch crystallization processes. The population balance model of aspirin crystallization process is first developed to describe the formation of crystals through nucleation and growth. Then, the PIRNN modeling scheme is introduced to integrate observational data and mechanistic models for the development of machine learning models. Subsequently, the PIRNN model that captures the dynamic behavior of the batch crystallization process is utilized in the design of model predictive control. Through open-loop and closed-loop simulations, it is demonstrated that the PIRNN models using less training data achieve prediction accuracy and closed-loop performance comparable to the purely data-driven model.

Keywords: Machine learning and data analytics in process control, Estimation and fault detection, Monitoring and performance assessment
Abstract: In industrial processes, deep learning has been widely used to solve the soft sensing problem. Multi-step-ahead prediction is one of the most challenging problems in the field of soft sensors. Recently, N-Beats has been proposed as a promising deep neural architecture for multi-step prediction, but it can only be used for univariate time series prediction, not for industrial soft sensor modeling. Inspired by N-Beats, a novel deep learning model, multivariate deep reconstruction neural network (MDRNN), is proposed for multivariate time series prediction in this work. MDRNN is designed on the basis of a doubly residual structure with a deep stack of fully-connect layers. MDRNN inherits the merits of N-Beats and incorporates the ``doubly residual stacking'' idea into the industrial soft sensor modeling to improve the prediction accuracy. To evaluate the feasibility and effectiveness of the proposed MDRNN, it is applied to the quality prediction task and validated with two real-world industrial processes. The application results demonstrated that the proposed MDRNN can achieve higher prediction accuracy compared to the existing N-Beats and Multilayer Perceptron (MLP)-based methods.

Keywords: Machine learning and data analytics in process control, Control system design, Process control applications
Abstract: This paper presents a machine learning application, to predict the formed nitrogenoxides (NOx) in thermal power plants, and used in the control loops of the selective catalytic reduction (SCR) process. To deal with the big plant operation data, data reduction methods are also described. The predictions are applied for the feed-forward (FF) control and show improvement of control performance in the simulation study.

Keywords: Distributed nonlinear control, Lagrangian and Hamiltonian systems, Control and estimation of wave equations and systems of elasticity
Abstract: Discrete control laws for flexible beam models obtained using structure-preserving discretization procedures are few in the existing literature so far and many control laws, including distributed ones, exist for continuous time models. Much of the effort has focussed on the discretization of the flexible structure and the fidelity of such models. The work in this paper examines the feasibility of a discretized version of a continuous time distributed control strategy, with a few approximations, to the discrete variational integrator of a flexible beam, in our case, a flexible single-link manipulator (FLM). The nonlinear control strategy is inspired by a potential shaping method in the literature based on the Cosserat continuum framework. We provide stability proof for this continuous-time control and approximate it to a discrete-time geometric framework. We use a geometrically exact beam model in which the manipulator's configuration space lies on an infinite-dimensional Lie group. The model-based discrete-time control uses a Lie Group Variational Integrator obtained by applying the variational principle on the Lagrangian of the FLM. Results show that the discrete closed-loop system stabilizes the FLM at setpoints during point-to-point rotational maneuvers. The controller demands fine time-sampling and we acknowledge the associated challenges for real-time implementation of the control scheme and intend to pursue it as our future work.

Keywords: Control in neuroscience, Constrained control, Numerical methods for optimal control
Abstract: We present a toolchain for solving path planning problems for concentric tube robots through obstacle fields. First, ellipsoidal sets representing the target area and obstacles are constructed from labelled point clouds. Then, the nonlinear and highly nonconvex optimal control problem is solved by introducing a homotopy on the obstacle positions where at one extreme of the parameter the obstacles are removed from the operating space, and at the other extreme they are located at their intended positions. We present a detailed example (with more than a thousand obstacles) from stereotactic neurosurgery with real-world data obtained from labelled MRI scans.

Keywords: Sampled-data control, Time-varying systems
Abstract: The paper studies the relation between a nonlinear time-varying flat discrete-time system and the corresponding linear time-varying systems which are obtained by a linearization along trajectories. It is motivated by the continuous-time case, where it is well-known that the linearization of flat systems along trajectories results in linear time-varying systems which are again flat. Since flatness implies controllability, this constitutes an important verifiable necessary condition for flatness. In the present contribution, it is shown that this is also true in the discrete-time case: We prove that the linearized system is again flat, and that a possible flat output is given by the linearization of a flat output of the nonlinear system. Analogously, the map that describes the parameterization of the system variables of the linear system by this flat output coincides with the linearization of the corresponding map of the nonlinear system. The results are illustrated by two examples.

Keywords: Sampled-data control, Digital implementation, Control under communication constraints (nonlinearity)
Abstract: Nonlinear controls designed in continuous time suffer from an increasing model mismatch and performance degradation when implemented in sampled-data control loops at relatively low sampling rates. The model mismatch is reduced when higher order numerical integration is used as modeling basis to describe the open-loop and the desired closed-loop dynamics. We recently presented a modular and minimally invasive way to implement continuously designed nonlinear controls with high accuracy in discrete time based on Gauss-Legendre collocation. In this paper, we show how the main drawback of the latter approach, highly discontinuous control signals, is removed by using piecewise cubic spline interpolation, or equivalently, Lobatto collocation. We present the rationale behind our approach, set up the nonlinear systems of equations for the required one-step predictions of the target dynamics, and discuss several relations between Hermite, Lobatto and Gauss collocation. On a benchmark nonlinear simulation example, we illustrate and discuss the performance of the approach with increasing sampling times.

Keywords: Sampled-data control, Passivity-based control, Digital implementation
Abstract: This paper aims at describing a synthesis procedure of discrete-time, energy-based regulators for continuous-time port-Hamiltonian systems. The methodology consists of three steps. The first twos deal with the definition of a discrete-time approximation of the plant to be successively employed in the development of the control law. Here, the focus is mainly on the last step, i.e. on how to interconnect digital controller and plant. The coupling is implemented via a zero-order hold and relies on the solution of an optimisation problem that determines the "best" and "minimal" correction to be applied to the nominal action to achieve the same performances obtained when the regulator is in closed-loop with the discrete-time model of the plant. This is the reference scenario used by the designer to develop and tune the control law. The procedure (time-discretisation, control design and coupling implementation) is illustrated in an example.

Keywords: Discontinuous control, Output regulation, Uncertain systems
Abstract: This paper studies the robust output regulation problem of uncertain single-input single-output (SISO) discrete-time linear systems. To reject the effect of a sinusoidal disturbance that only the frequency information is known prior, a novel strategy based on adaptive feedforward control (AFC) is developed. Compared with existing regulators for uncertain discrete-time systems, neither the knowledge on the sign of the real part or the imaginary part of the transfer function at the frequency of interest (the so-called strictly positive real (SPR)-like condition), nor persistence of excitation condition is required in this approach. Stability of the closed-loop system is rigorously analyzed using small-gain theorem and Lyapunov-based stability theory. Essentially, the proposed scheme guarantees that all signals of closed-loop system are bounded while the output of system asymptotically converges to zero, which is demonstrated by a numerical example.

Keywords: Fault detection and diagnosis, Machine learning, Time series modelling
Abstract: Data-driven modeling and machine learning have received a lot of attention in fault diagnosis and system monitoring research. Since faults are rare events, conventional multi-class classification is complicated by incomplete training data and unknown faults. One solution is anomaly classification which can be used to detect abnormal behavior when only training data from the nominal operation is available. However, data-driven fault isolation is still a non-trivial task when training data is not representative of nominal and faulty behavior. In this work, the importance of redundancy for a set of known variables that are fed to a data-driven anomaly classification is discussed. It is shown that residual-based anomaly detection can be used to reject the nominal class which is not possible with one-class classifiers, such as one-class support vector machines. Based on these results, it is also discussed how data-driven residuals can be integrated with model-based fault isolation logic.

Keywords: Machine learning, Fault detection and diagnosis, Nonlinear system identification
Abstract: Data-driven modeling and machine learning are widely used to model the behavior of dynamic systems. One application is the residual evaluation of technical systems where model predictions are compared with measurement data to create residuals for fault diagnosis applications. While recurrent neural network models have been shown capable of modeling complex non-linear dynamic systems, they are limited to fixed steps discrete-time simulation. Modeling using neural ordinary differential equations, however, make it possible to evaluate the state variables at specific times, compute gradients when training the model and use standard numerical solvers to explicitly model the underlying dynamic of the time-series data. Here, the effect of solver selection on the performance of neural ordinary differential equation residuals during training and evaluation is investigated. The paper includes a case study of a heavy-duty truck's after-treatment system to highlight the potential of these techniques for improving fault diagnosis performance.

Keywords: Fault detection and diagnosis, Machine learning, Grey box modelling
Abstract: Flotation in the mining industry is of vital importance for obtaining the right quality of product with efficiency and represents a critical process where possible failures must be monitored at all times. In this paper, complete fault detection and isolation system (FDI) based on the Neural Ordinary Differential Equations (NODE) framework is proposed; the NODE is employed to represent the dynamics of the studied plant based on the measured variables and inputs. Then, a classifier can be used to identify the faults based on the projections of the derivatives or local vector field generated by the NODE using the estimations and actual measurements. The proposed approach is applied to a controlled mining flotation process that has perturbations. The solution is compared with other known machine learning techniques showing better performance metrics. Moreover, it is demonstrated with t-SNE representation that features generated from the NODE model improve the classification.

Keywords: Fault detection and diagnosis
Abstract: To study large-scale systems, the usual formal mathematical tools based on analytical models derived from differential and algebraic equations cannot in general be used due to the high complexity of the system structure. The underlying physical laws are usually unknown in analytical form or are too complicated for calculations. To overcome this complexity, methods based on what is called artificial intelligence (AI) and exploiting system data can be an efficient alternative to model-based methods. In this article, we propose a method based on the transfer entropy analysis to identify the causal relationships between the measured variables of the process in order to have a graphical representation model of the system. This graphical model can be used in root cause and hazard propagation analysis. For system performances evaluation and Fault Diagnosis, the large-scale functionality due to the large number of variables is solved by limiting the data to be processed in the modeling phase to key performance indicators (KPIs) and loop performance indicators (LPIs) that really have an impact on the behavior of the process. The engineering methodology is organized in two stages, respectively Topological analysis defining the technical failure and the equipment levels and Root cause analysis defining the relations between the defects and the causes. For fault detection, local residuals are analyzed online using specific statistical hypothesis tests applied to KPIs and LPIs. A case study based on a three-store heating system is presented to illustrate the application of the proposed methods.

Keywords: Machine learning, Fault detection and diagnosis, Time series modelling
Abstract: In recent years, 3D printing has seen a stellar rise despite its inability to deliver constant quality goods. This article presents a machine learning experiment that results in a model performing fault prediction, in the sense of forecasting the fault, on the printed parts so that printer parameters can be corrected before the faults appear. This model is able to predict faults in real-time during printing, even in the case of multiple faults. It relies on multiple sensors gathering time-series data during printing, a pre-processing of these data to extract the most relevant features and several machine learning algorithms, each suited and tuned to predict at best each fault. A benchmark for testing and tuning the different algorithms is presented. The resulting model has been implemented on a plastic delta 3D printer and tested for the prediction of eight different faults. The best performing model is a random forest, but decision trees are almost as good while explaining what causes the fault.

Keywords: Stochastic adaptive control, Fault detection and diagnosis
Abstract: Multiple model adaptive control (MMAC) provides robust control guarantees for a range of plant configurations given a set of models that correspond to known operating conditions (modes). However, these guarantees may not hold in the face of unknown system configurations. This paper focuses on the first step of control under novel modes: the task of novel configuration detection. We analyze a change-point detection strategy for unknown system configurations using input/output data as input. We characterize the performance of the change-point detection strategy based on whether the performance loss is bounded or unbounded. We experimentally validate our algorithms using a quadcopter trajectory-tracking benchmark, comparing our approach to a Bayesian change-point detection strategy.

Keywords: Networks of robots and intelligent sensors, Multi-vehicle systems, Decentralized control and systems
Abstract: This paper proposes a distributed persistent coverage strategy for environments with important regions. The strategy involves using quadrotors at varying altitudes to acquire high-resolution information from these regions. The information is then transmitted through distributed communication protocols that use local coverage maps to retain periodic and high-quality data. To resolve the trade-off between high-resolution data from low altitudes and wider coverage from high altitudes, cost functions based on distance, coverage, and sensor resolution are used to determine three-dimensional waypoints. Further, Voronoi tessellations are used to aid in collision avoidance and efficient computation. To generate feasible trajectories, the quadrotor's dynamics are considered, along with error functions that use state information and desired waypoints. Simulation results demonstrate the efficacy of the proposed approach for multi-agent systems under quadrotor dynamics.

Keywords: Sensing, Trajectory tracking and path following, Autonomous mobile robots
Abstract: This paper presents a minimum sensing strategy for a path-following task where a mobile robot is required to satisfy a prescribed localization accuracy by measuring known landmarks. Specifically, we design a sensing strategy that confines the covariance of a robot to the pre-required locational accuracy along a path by controlling the degree of attention to each landmark. We first present a novel discrete-time control barrier function (DCBF) that confines the covariance of the robot inside of the pre-planned locational requirement. We then integrate the proposed DCBF into the optimization problem, which is designed to allocate the attention of a robot to each landmark so that the overall sensing effort is minimized. In the proposed optimization problem, we evaluate the sensing cost as the minimum information gain, namely the minimum number of bits that must be included in the sensor data, and formulate how the specified attention level affects the estimation of the robot state. Finally, we demonstrate the effectiveness of the proposed method in simulation studies.

Keywords: Sensing, Health monitoring and diagnosis, Map building
Abstract: The measurement of CO2 concentrations is becoming increasingly important in order to avoid poorly ventilated enclosed spaces. This paper introduces the background law, theory, and technology regarding indoor CO2 concentration. We developed a novel palm-sized measuring device, Pocket CO2 Sensor. This IoT-type sensor-based risk visualization and actual ventilation countermeasures will be presented as a case study. A new era of networked air-conditioning management will be coming by visualizing and sharing risk information.

Keywords: Automotive sensors and actuators
Abstract: In the fields of agriculture and disaster prevention, it is necessary to monitor soil pH and moisture conditions. Our group has been studied a multimodal sensor on a small Si chip integrated with a pH, water content, and temperature sensors. We succeeded to fabricate the chip using Si-LSI technology. We succeeded to measure soil pH with low water content and water content of expressway slope.

Keywords: Information displays/system, Man-machine interfaces, Robot navigation, programming and vision
Abstract: Homography transformation, or planar projective transformation, between camera and projector images, which can be estimated by observing four or more pairs of corresponding points between two images, plays a crucial role in dynamic projection mapping with an uncalibrated projector-camera pair. This study discusses a case of dynamic projection mapping onto a moving textured surface where the point correspondence is provided through an actively controlled (not static) marker pattern projection by the projector along with the movement of the surface. Qualitative advantages are discussed in terms of system implementation, the image processing algorithm, and artifact perception by human observers. In addition, a quantitative analysis of alignment errors in the projection mapping is presented to show the effectiveness of the active control of marker patterns.

Keywords: Localization, Robot navigation, programming and vision, Map building
Abstract: The article presents a monocular visual localization system based on fiducial markers for autonomous docking in industrial applications. The design has been tailored for direct use with the docking of a mobile logistic robot under storage racks that requires high robustness, usable accuracy, and easy maintenance at a very low cost. The elaborated solution suggests an innovative calibration tool for determining the positions of used fiducial markers at the docking location, which is based on structure-from-motion reconstruction. Moreover, the paper introduces a localization system calculating the position of the robot relative to the docking location using the Perspective-n-Point Problem solver for the detected markers. The presented approach has been verified for precision and robustness limits in diverse work conditions and quantitatively evaluated in experiments.

Keywords: Advanced planning and scheduling, Digital twins for manufacturing, AI-based enterprise systems
Abstract: Digital manufacturing tools aim to provide intelligent solutions that can support manufacturing industry to adapt to the volatile operational environment. The successful implementation of such tools highly depends on the capabilities of the digital frameworks or platforms they are deployed upon as well as the quality of their intelligence. The objective of this work is to develop and discuss a framework for training and deploying Machine Learning (ML) agents for production scheduling with the augmentation of Digital Twin (DT) technologies. Two types of ML production scheduling agents have been developed and integrated with the DT framework: a Deep Learning agent and a Deep Reinforcement Learning agent. In order to increase interoperability, Asset Administration Shell Industry4.0 standard has been utilized for the integration and deployment of the proposed DT framework into industrial practice. The proposed framework is tested and validated upon an industrial case study from the bicycles’ production industry.

Keywords: Manufacturing plant control, Intelligent manufacturing systems, Smart factory
Abstract: Modular production systems provide an important basis to react more flexibly to changing and more individual production requirements. A challenge hereby is an appropriate production planning and scheduling that considers the capabilities and the degrees of freedom of these systems, which can be algorithmically optimized for their relevance. A great importance is also given to the aspect of energy-efficient production, which can be included in the planning as well as in its execution on the modular level. The main objective of the paper is to show the relevance of the adaptability of production plans within modular production systems from an energetic point of view, with special focus on electrical energy. It is considered on how modular production systems can be modeled, based on the current state of the art and from the perspective of the capability- and skill-based approach for energy-aware skill usage. Furthermore, it is examined how pre-optimized scheduling for production systems can be adapted by production modules. For the energy-related optimization an extensive-form game model is provided. The scheduler and the production modules are considered as players and strategies of the modules are selected based on their modeled capabilities by using their best response under energy-related aspects. The results of a simulation with prototypical implementation for energy-optimized adaptation, based on extensive-form games, and possible application for unforeseen event handling are presented and discussed.

Keywords: Manufacturing plant control, Intelligent manufacturing systems, Industry 4.0 
Abstract: The current volatility on the global market presents new challenges in production system engineering and control. On the one hand, packaging and packaging production systems must support increasing production volumes. On the other hand, market volatility requires more flexible production that is fragmented into smaller batches, which causes increased time, effort, and overall workload implied by changeovers between different production setups. Typically, it is not feasible to build a green-field generation of production sites that conforms to all design principles of Industry 4.0. Therefore, the addressed approach deals with a "smartification" of legacy hardware components, bridging multiple levels of the traditional automation pyramid, improving human-machine collaboration, and transforming human experience into a common shared knowledge base. In more detail, we are applying unsupervised learning methods to generalize individual machine settings into reusable and automatically applicable control setups and strategies. The core components of the addressed packing lines are counting machines that deliver the required portions of elements. The developed Smart Counting Machines retain current interfaces that are enhanced with new enhancements providing OPC UA communication and modular control layer. The realized solution has been deployed and tested in a real packing line of our industrial partner. The lessons learned in the presented mitigate barriers on the journey from the proof-of-concept phase to the full production deployment.

Keywords: Multi-agent systems applied to industrial systems, Smart factory, Industry 4.0 
Abstract: Robotic cells in industrial automation use different programming and description languages, which are typically almost hard-wired in the program solutions and do not allow a goal product to be changed easily. Using a unification description language allows focusing on the solution itself. An independent description of capabilities and intentions offers a way to allow for changing the goal and also for distributing the process to different locations. Moreover, based on the description, a knowledge model can be created to check if a solution exists for the defined goal and resource capabilities. If the knowledge model is implemented in a database, the check can be performed in a very efficient way, which allows it to be used in real-life production scenarios. A reasoner tries to reach the solution-defined goal based on an initial state and actions following the predicates' rules. Using the database to search only for the first possible solution, it can be checked if the defined domain model can be realized. The check can be done automatically and much faster than using a planner. In an implementation in our Testbed for Industry 4.0 at CIIRC/CTU Prague, we employ a scalable system based on a PDDL description and an automated translation to TypeDB to efficiently compute production plans for changing goals, tools and resources.

Keywords: Model driven engineering of control systems
Abstract: The automation of plants in the process industry, especially of modular process plants, is shifting towards a more decentralized and service-based paradigm. Required and offered services need to be described in such a way that they can be either easily described by plant operators and module vendors. This two-sided problem requires a semantic description of needed functionalities from intentions and abstraction of existing services onto offered functionalities. In this contribution, the authors present two algorithms for the modeling and derivation of functionalities from intentions and services in modular automation. The algorithms are based on the idea of functional reasoning which support the explanation and derivation of functions in a system. By using various ontologies, the approach builds upon semantic descriptions and exploits query- and rule-based languages to support the retrieval of functionalities.

Keywords: Biomedical system modeling, simulation and visualization, Pharmacokinetics and drug delivery
Abstract: Fast simulation of linear time-invariant (LTI) pharmacokinetic (PK) models is crucial to mixed-effect modeling techniques, used extensively in pharmacological research and development. The by far most common LTI PK models are particularly structured compartmental systems with one, two or three compartments. Here we develop and demonstrate very efficient, and down to machine precision exact, simulators for those structures. Our proposed method is benchmarked against state-of-the art software for simulation of linear systems, using a clinically relevant data set.

Keywords: Medical imaging and processing, Biomedical and medical image processing and systems
Abstract: As recent advances in AI are causing the decline of conventional diagnostic methods, the realization of end-to-end diagnosis is fast approaching. Ultrasound image segmentation is an important step in the diagnostic process. An accurate and robust segmentation model accelerates the process and reduces the burden of sonographers. In contrast to previous research, we take two inherent features of US images into consideration: (1) different organs and tissues vary in spatial sizes, (2) the anatomical structures inside human body form a relatively constant spatial relationship. Based on those two ideas, we propose a new image segmentation model combining Feature Pyramid Network (FPN) and Spatial Recurrent Neural Network (SRNN). We discuss why we use FPN to extract anatomical structures of different scales and how SRNN is implemented to extract the spatial context features in abdominal ultrasound images.

Keywords: Medical imaging and processing, Biomedical and medical image processing and systems, Modeling and identification
Abstract: The degree of progression of a malignant tumor in the gastrointestinal tract is determined primarily by an endoscopist based on endoscopic images. The most important factor in determining the degree of progression is the tumor size. In this paper, we propose a method to three-dimensionally model of the stomach lining from endoscopic images and to measure the tumor by using the data. While conventional methods use a reflection model wherein the endoscope has a single light source, the proposed method constructs a new reflection model that leverages the characteristics of the two light sources. Finally, we demonstrate the effectiveness of the proposed method through several experiments.

Keywords: Control of physiological and clinical variables, Pharmacokinetics and drug delivery
Abstract: This paper investigates the use of different pharmacokinetic/parmacodynamic models for the design of a PID-based control structure for total intravenous anesthesia. In particular, a PID controller tuned by exploiting the Schnider model for propofol administration (which has been developed for adults) in order to ensure the required robustness and performance is evaluated by considering patients modelled by using the recently developed Eleveld model. The latter is more general and, in principle, takes into account children, adults and elderly people. Simulation results show that clinically acceptable performance are obtained, disregarding the adopted model, for children and adults. On the contrary, the use of the Eleveld model to tune the controller for elderly people, for which a significant delay emerges, requires further investigation.

Keywords: Healthcare management, disease control, critical care, Identification and validation, Physiological Model
Abstract: Neonatal hyperglycaemia is common for infants in critical care, and can lead to increases in mortality and morbidity. Subcutaneous insulin delivery systems offer an easy way to control blood glucose (BG) to more normal levels. However, research on subcutaneous insulin models are lacking in this cohort. This paper presents a model combining validated NICU models with adult subcutaneous models for glycaemic control. Clinical data from 12 very/extremely pre-mature infants was collected for an average study duration of 10.1 days. Glucose, interstitial and plasma insulin, as well as subcutaneous and local insulin were modelled, and insulin sensitivity profiles were identified for each patient. Modelling error was low, where the cohort median [IQR] for mean percentage error was 0.8 [0.3-3.4] %. The model was able to capture glucose and insulin activity accurately, but further research should be undertaken to validate the model and its parameters, to explore the potential of subcutaneous insulin delivery under a model-based protocol.

Keywords: Decision support and control, Intensive and chronic care or treatment, Control of physiological and clinical variables
Abstract: Insulin sensitivity is a key physiological parameter in the human metabolic system. Insulin dosing of insulin-dependent intensive care patients can be guided by model-based glycemic control protocol using insulin sensitivity prediction to calculate the treatment outcomes and find the optimal treatment option. Quantile regression is an artificial intelligence method found to be efficient in the prediction of processes with stochastic behaviour. Using the patent data set comprising data from 3 clinical ICU cohorts and including 820 treatment episodes of 606 patients and 68,631 hours of treatment, an insulin sensitivity prediction model is created using quantile regression. Besides the standard prediction metrics, Success rate, Interval ratio, and I-Score are applied to evaluate the efficacy of the prediction in the context of clinical requirements. Several activation functions, network topologies, and epoch numbers are tested to optimize the method’s hyper-parameters. Based on the results presented, the quantile regression-based insulin sensitivity prediction methods proposed in this paper are found to be accurate enough for application in glycaemic control protocols controlling the insulin dosing of intensive care patients. The accuracy of the artificial intelligence methods is very similar to the currently used stochastic model-based prediction. Thus, it is worth to be evaluated by in-silico simulation trials and, in case of successful validation being tested by clinical trials.

Keywords: Reinforcement learning and deep learning in control, Multi-agent systems applied to industrial systems, Industrial applications of optimal control
Abstract: Efficiently solving path planning problems for a large number of robots is critical to the successful operation of modern warehouses. The existing approaches adopt classical shortest path algorithms to plan in environments whose cells are associated with both space and time in order to avoid collision between robots. In this work, we achieve the same goal by means of simulation in a smaller static environment. Built upon the new framework introduced in (Bertsekas, 2021a), we propose multiagent rollout with reshuffling algorithm, and apply it to address the warehouse robots path planning problem. The proposed scheme has a solid theoretical guarantee and exhibits consistent performance in our numerical studies. Moreover, it inherits from the generic rollout methods the ability to adapt to a changing environment by online replanning, which we demonstrate through examples where some robots malfunction.

Keywords: Multi agent systems, Reinforcement learning and deep learning in control
Abstract: Many multiagent systems in the real world include multiple types of agents with different abilities and functionality. Such heterogeneous multiagent systems have significant practical advantages. However, they also come with challenges compared with homogeneous systems for multiagent reinforcement learning, such as the non-stationary problem and the policy version iteration issue. This work proposes a general-purpose reinforcement learning algorithm named Heterogeneous League Training (HLT) to address heterogeneous multiagent problems. HLT keeps track of a pool of policies that agents have explored during training, gathering a league of heterogeneous policies to facilitate future policy optimization. Moreover, a hyper-network is introduced to increase the diversity of agent behaviors when collaborating with teammates having different levels of cooperation skills. We use heterogeneous benchmark tasks to demonstrate that (1) HLT promotes the success rate in cooperative heterogeneous tasks; (2) HLT is an effective approach to solving the policy version iteration problem; (3) HLT provides a practical way to assess the difficulty of learning each role in a heterogeneous team.

Keywords: Telecommunication-based automation systems
Abstract: Networked control systems using wireless links to convey information among sensors, controllers, and actuators greatly benefit from having an accurate estimate of the communication channel condition. To this end, the finite-state Markov channel abstraction allows for reliable channel state estimation. This paper develops a Markov jump linear system representation for wireless networked control with intermittent channel state observation, message losses, and generalized hold-input dropout compensation. Furthermore, it exploits the emerging structural properties of the system to solve the finite-horizon linear quadratic regulation problem efficiently.

Keywords: Tele-robotics, Soft computing in control, Evolutionary algorithms in control and identification
Abstract: In this paper a novel improved multi-objective genetic algorithm (R-NSGA-Ⅱ) is proposed to solve the problem of dual-robot arc welding path planning. We strive to find a low-cost, fast and more efficient solution for large complex components. Firstly, a multi-objective optimization model of path planning is established by considering various variables and constraints in the actual welding process. Then, the R-NSGA-Ⅱ introduces a recombination mechanism to replace the mutation operation of NSGA-Ⅱ; this can increase the diversity of individuals and populations, and find the global optimal solution with greater probability. Finally, in order to verify feasibility and effectiveness of the proposed algorithm, it is used to plan some typical welding seams of a large complex component, the NSGA-Ⅱ and OMOPSO are used for comparison. The simulation demonstrates that R-NSGA-Ⅱ can obtain better Pareto front than the other two algorithms. Particularly, the R-NSGA-Ⅱ reduces the waiting time by 42.39% and the no-load distance by 10.50% compared with NSGA-Ⅱ.

Keywords: Telecommunication-based automation systems, Remote and distributed control, Traffic control systems
Abstract: Software Defined Networking (SDN) is a networking architecture within the control is centralized through a software-based controller. Hence, the controller represents a single point of attack which makes the controller a preferred target in case of attack. In previous work, an observer has been introduced in order to detect any anomalies in the network control (in particular, in the data planes which are set up). The detection method is based on the assumption that the observer captures all the packets sent and received by the controller. This paper introduces an extension of the detection approach to improve the robustness in case of partial or parsimonious observation, more specifically when the complete data plane cannot be observed. The evaluation of the consistency of the data plane is adapted by definition of necessary condition which permits to conclude that the data plane is inconsistent. In the absence of evidence, a function which permits to compute the level of confidence in the consistency of the data plane is introduced. The efficiency and the limitations of such method is discussed on a case study.

Keywords: Tele-maintenance, Tele-robotics, Remote sensor data acquisition
Abstract: This paper focuses on the train under-vehicle inspection robot equipped with 3D solid-state LiDAR and laser distance sensor. Dedicated 3D laser Simultaneous Localization and Mapping (SLAM) framework and motion control algorithm are proposed for the robot operating in the hazardous environment of the inspection pit. The constrained working environment includes repetitive and monotonous spatial features, which makes autonomous localization and navigation task a challenging issue. The proposed SLAM localization framework has a complete architecture that includes frontend laser odometry and backend loop closure detection. The method optimized for 3D solid-state LiDAR is utilized for feature extraction. The pose optimization problem is initialized with the motion estimation provided by wheel odometry. The wheel odometry interpolation is utilized to increase the pose update frequency. Moreover, the features of the vehicle wheels in the measurements of laser distance sensor are utilized for loop closure detection. The proposed motion control algorithm extracts the features of sidewalls for feedback of angular velocity closed-loop control. The experimental results from simulation and real-world show that the robot can be driven accurately along a straight trajectory in the inspection pit with a maximum integrated localization and navigation error of less than 0.015m in movement over 200m, which demonstrates the effectiveness of the proposed algorithms.

Keywords: Machine learning in modelling, prediction, control and automation, Time series modelling, Learning and adaptation in autonomous vehicles
Abstract: Road grade is an important parameter for application of modern vehicle intelligent control technology. At present, the estimation methods of road grade are mainly based on the vehicle longitudinal dynamics or sensor information. However, for practical applications, relying only on the recognition of road grade, without map information, will be a post estimation, which is not conducive to the implementation of intelligent control strategies. To meet the needs of predictive intelligent control algorithms for vehicle, it is necessary to predict future road grade changes. This study considers the road grade variation experienced by the vehicle as a time series and proposes a prediction method that combines a convolutional neural network (CNN) and a long short-term memory neural network (LSTM). The road grade information collected by the vehicle driving sensors is used to train the designed network to predict the future road grade and the prediction model is compared with LSTM prediction model. The results of experimental data analysis show that the method designed in this study is superior to LSTM prediction model, and on test set, the MAPE (Mean Absolute Percentage Error) of the road grade prediction after 5 s is 11.6 %, which is 15.1 % lower than the LSTM prediction result. The generalization ability and computational time of the model are also compared, and the CNN-LSTM demonstrated better performance.

Keywords: Machine learning in modelling, prediction, control and automation, Neural networks, Engine modelling and control
Abstract: The increasing demand of lowering the emissions of the combustion engines has led to the development of more complex engine systems. This paper presents artificial neural network (ANN) based models for estimating nitrogen oxide (NOx) and carbon dioxide (CO2) emissions from in-cylinder pressure of a maritime diesel engine. The architecture of the models is that of Multi-Layer Perceptron (MLP) and Radial Basis Function (RBF) network. The data utilized to train and test the models are obtained from a four-cylinder marine engine. The inputs of the models are chosen as the first principal components of the in-cylinder pressure and engine parameters with highest correlation to aforementioned greenhouse gases. Generalization is performed on the models during the training to avoid overfitting. The estimation result of each model is then compared. Additionally, contribution of each cylinder to the production of emissions is investigated. Results indicate that MLP has a higher accuracy in estimating both NOx and CO2 compared to RBF network. The emission levels of each cylinder for both NOx and CO2 are mostly even due to the nature of the conventional diesel engine.

Keywords: Machine learning in modelling, prediction, control and automation
Abstract: The European XFEL is one of the largest linear particle accelerator in the world used to generate extremely intense and ultra short X-Ray flashes to study ultra fast, time resolved chemical processes. In order to improve the quality of the observations, the laser-based optical synchronization system is optimized by tuning PI controllers using a safe Bayesian optimization approach. Since machine time on the European XFEL is very expensive, the algorithm needs to find the optimal parameters as fast as possible. In this contribution, we present a safe Bayesian optimization algorithm which, while guaranteeing safety, shows significantly improved convergence speed and noise robustness. Application and comparison results are presented in simulation for the optical synchronization system of the European XFEL and an experimental demonstration is performed for a laboratory synchronization system.

Keywords: Machine learning in modelling, prediction, control and automation, Sliding mode control, UAVs
Abstract: This paper proposes the approach of employing state space least squares support vector machine (SS LS-SVM) as a disturbance observer in the sliding mode control of a quadrotor. SS LS-SVM, which was recently introduced by the authors, is adopted for the disturbance estimation task in this study. A quadrotor type unmanned aerial vehicle is considered as the system of interest to apply and assess the performance of SS LS-SVM as a disturbance observer. Quadrotor continuous time mathematical model is taken into account in a standard integrator based on Euler discritization. Both parametric uncertainties and external disturbances are lumped in a disturbance term and added to the nominal model. That term is approximated by SS LS-SVM in an output error prediction context by minimizing the state estimation error via gradient descent as the training method. The proposed disturbance observer works in collaboration with a standard nonlinear observer. It is only necessary for estimating the system states using the measured system output while SS LS-SVM performs the estimation of disturbance. SS LS-SVM enables placement of a native LS-SVM directly in a state equation. Simulation results indicates the significant performance of closed loop disturbance estimation by the SS LS-SVM disturbance observer and based on that, robustness of the employed control method is empowered.

Keywords: Machine learning in modelling, prediction, control and automation, Data-based control, Autonomous systems
Abstract: This paper presents a method to design neural network controllers based on imitation learning and with tunable stability guarantees through multi-objective optimization. Stability margins are derived from analyzing state-space neural networks based on the representation of nonlinear activation functions by linear parameter varying models. The controller training is formulated as a multi-objective problem whose solutions yield a set of the best trade-offs in terms of minimal imitation error and maximal stability margins. The proposed approach is illustrated in the synthesis by imitation of an aircraft neural autopilot using a flight simulator.

Keywords: Machine learning in modelling, prediction, control and automation, Smart energy grids, Machine learning
Abstract: When developing load forecasting models a common strategy is to train a range of different models in order to determine the best model for prediction. Rather that retaining only the best model, employing stacked ensemble methods to combine the predictions from all the models can often yield superior prediction performance to any of the constituent models. In this paper we explore four different approaches for generating such stacked ensemble predictions: (1) simple heuristic rules; (2) conformal learning inspired model confidence weighting approaches; (3) optimal model classifiers, and; (4) non-linear regression based approaches. Short-term load forecasting for the Northern Ireland and New York State power systems are used as case studies. Our results show that the non-linear regression based stacked ensemble model yields the most consistent performance across the case studies, achieving reductions in mean absolute prediction error (MAPE) of between 10% and 14% relative to the best performing individual models.

Keywords: Large scale optimization problems, Probabilistic robustness, Logistics in manufacturing
Abstract: The demand for resilient logistics networks has increased because of recent disasters. With regard to optimization problems, entropy regularization is a powerful tool for the diversification of a solution. In this study, we devised a method for designing a resilient logistics network based on entropy regularization. Moreover, we developed a method for analytical resilience criteria to reduce the ambiguity of resilience. First, we modeled the logistics network, including factories, distribution bases, and sales outlets in an efficient framework using entropy regularization. Next, we formulated a resilience criterion based on probabilistic cost and Kullback–Leibler divergence. Finally, the proposed method was implemented using a simple logistics network, and the resilience of three logistics plans designed using entropy regularization was demonstrated. Consequently, it is confirmed that our criteria can be used to evaluate resilience to logistics disruptions.

Keywords: Linear parameter-varying systems, Robust control, Aerospace
Abstract: This paper presents a systematic linear parameter varying (LPV) control approach for the 3-axis attitude control of an Earth-observation satellite in a sun-synchronous orbit. The dynamics of the satellite depend on the orientation of the solar array, which completes a full rotation every orbit, thus it is used as a scheduling parameter in the design. The satellite has two additional flexible appendages; these are 2 scatterometers. The control objective is to precisely track a given reference attitude using reaction wheels, while rejecting external torque disturbances and sensor noise. The design follows a mixed-sensitivity approach, applying a recently introduced weighting scheme. It allows traceable and effective controller tuning by using a low number of physically interpretable weights. The controller is synthesised by solving the induced L2-norm of the closed-loop interconnection of the controller and weighted plant. Scheduling with the solar array orientation leads to an LPV notching behaviour in the controller that effectively mitigates the effects of the array’s most prominent flexible modes. This behaviour enables increased performance, when compared to a linear time invariant controller, while maintaining robustness. The pointing performance of the synthesised controller over the complete satellite lifecycle is verified using the European Space Agency’s standards for spacecraft attitude control.

Keywords: Real-time optimal control, Convex optimization, Guidance navigation and control
Abstract: We introduce a multi-phase rocket landing guidance framework that can handle nonlinear dynamics and does not mandate any additional mixed-integer or nonconvex constraints to handle discrete temporal events/switching. To achieve this, we first introduce sequential conic optimization (SeCO), a new paradigm for solving nonconvex optimal control problems that is entirely devoid of matrix factorizations and inversions. This framework combines sequential convex programming (SCP) and first-order conic optimization and can solve unified multi-phase trajectory optimization problems in real-time. The novel features of this framework are: (1) time-interval dilation, which enables multi-phase trajectory optimization with free-transition-time; (2) single-crossing compound state-triggered constraints, which are entirely convex if the trigger and constraint conditions are convex; (3) virtual state, which is a new approach to handling artificial infeasibility in SCP methods that preserves the shapes of the constraint sets; and, (4) the use of the proportional-integral projected gradient method (PIPG), a high-performance first-order conic optimization solver, in tandem with the penalized trust region (PTR) SCP algorithm. We demonstrate the efficacy and real-time capability of SeCO by solving a relevant multi-phase rocket landing guidance problem with nonlinear dynamics and convex constraints only, and observe that our solver is 2.7 times faster than a state-of-the-art convex optimization solver.

Keywords: Industrial applications of optimal control, Non-smooth and discontinuous optimal control
Abstract: This paper addresses the trade-off between time- and energy-efficiency for the problem of loading and unloading a ship. Container height constraints and energy consumption and regeneration are dealt with. We build upon a previous work that introduced a coordinate system suitable to deal with container avoidance constraints and incorporate the energy related modeling. In addition to changing the coordinate system, standard epigraph reformulations result in an optimal control problem with improved numerical properties. The trade-off is dealt with through the use of weighting of the total time and energy consumption in the cost function. An illustrative example is provided, demonstrating that the energy consumption can be substantially reduced while retaining approximately the same loading time.

Keywords: Optimal control theory, UAVs, Tracking
Abstract: This paper proposes a linear quadratic gaussian control for trajectory tracking of a quadrotor UAV. It involves implementing an optimal linear quadratic regulator control with integral action in an inner-outer loop control architecture. The full-state multi-rate extended Kalman filter generates the feedback required for the optimal control. Biases in gyroscope and accelerometer measurements are also incorporated into the Kalman filter to avoid degradation of the closed-loop response and to provide accurate state estimates. The proposed control architecture is tested on an experimental test bed consisting of a quadrotor UAV platform equipped with a Gumstix DuoVero Zephyr microcontroller. The onboard inertial measurement unit, altimeter, and motion capture system provides the necessary measurements. The recorded results validate the performance of the proposed control scheme with improved state feedback generated through the extended Kalman filter.

Keywords: Industrial applications of optimal control, Nonlinear and optimal marine system control, Control of constrained systems
Abstract: In this paper, we investigate autonomous vessel operation in/near a port and present a novel hierarchical control architecture that combines model predictive control (MPC) and control barrier function (CBF). Our architecture is composed of two layers, navigation and control, and switches control schemes as the ideal vessel operations vary depending on the vessel locations. Specifically, we divide the operation into three phases, namely approaching phase, breakwater passing phase, and docking phase. In all phases, we employ the CBF-based online optimization with a vessel kinematic model at a lower layer to certify safety even in the presence of the prediction error in the navigation layer. The navigation layer is designed based on MPC so as to smoothen the collision avoidance behavior and to reflect various specifications stemming from laws. The present control architecture is demonstrated through various simulations including the one with real data of vessels in Tokyo Bay.

Keywords: Decentralized control, Large scale optimization problems, Predictive control
Abstract: Coalitional control partitions a system into multiple clusters or coalitions that solve independent local subproblems in parallel. This paper presents a two-layer coalitional model predictive control approach for regulation in constrained-coupled subsystems. We formulate a resource allocation mechanism to distribute the coupled constraint so that the global control problem can be solved in a decentralized manner, guaranteeing the satisfaction of the common constraint. In particular, a top layer will calculate the system's partition according to a given criterion and supervise the shared resource allocation. In turn, the lower control layer will calculate the local optimization problems for every coalition in a decentralized fashion, according to the available shared resource determined by the upper layer. This strategy is applied to regulate the outlet temperature of parabolic-trough solar collector fields, which are composed of a set of loops that remain coupled through a global shared resource constraint.

Keywords: Robust control (linear case), Model predictive and optimization-based control, Constrained control
Abstract: In this paper, we derive set constraints for a reduced order model and augment them into a model predictive control (MPC) scheme to ensure safe operation of the largescale ensemble system. For the control feedback, only the aggregated information of the whole system is required. For the constraint satisfaction, we consider an adaptive tube formulation to characterize the deviation between the reduced order model and the ensemble system. Employing the robust control invariant set, we ensure recursive feasibility and initial feasibility under an easily verifiable condition.

Keywords: Predictive control, Nonlinear predictive control, Multi-agent systems
Abstract: We consider multi-agent systems with heterogeneous, nonlinear agents subject to individual constraints that want to achieve a periodic, dynamic cooperative control goal which can be characterised by a set and a suitable cost. We propose a sequential distributed model predictive control (MPC) scheme in which agents sequentially solve an individual optimisation problem to track an artificial periodic output trajectory. The optimisation problems are coupled through these artificial periodic output trajectories, which are communicated and penalised using the cost that characterises the cooperative goal. The agents communicate only their artificial trajectories and only once per time step. We show that under suitable assumptions, the agents can incrementally move their artificial output trajectories towards the cooperative goal, and, hence, their closed-loop output trajectories asymptotically achieve it. We illustrate the scheme with a simulation example.

Keywords: Predictive control, Nonlinear predictive control, Digital implementation
Abstract: We present IMPACT, a flexible toolchain for nonlinear model predictive control (NMPC) specification with automatic code generation capabilities. The toolchain reduces the engineering complexity of NMPC implementations by providing the user with an easy-to-use application programming interface, and with the flexibility of using multiple state-of-the-art tools and numerical optimization solvers for rapid prototyping of NMPC solutions. IMPACT is written in Python, users can call it from Python and MATLAB, and the generated NMPC solvers can be directly executed from C, Python, MATLAB and Simulink. An application example is presented involving problem specification and deployment on embedded hardware using Simulink, showing the effectiveness and applicability of IMPACT for NMPC-based solutions.

Keywords: Predictive control, Optimal control theory
Abstract: For a receding-horizon controller with a known system and with an approximate terminal value function, it is well-known that increasing the prediction horizon can improve its control performance. However, when the prediction model is inexact, a larger prediction horizon also causes propagation and accumulation of the prediction error. In this work, we aim to analyze the effect of the above trade-off between the modeling error, the terminal value function error, and the prediction horizon on the performance of a nominal receding-horizon linear quadratic (LQ) controller. By developing a novel perturbation result of the Riccati difference equation, a performance upper bound is obtained and suggests that for many cases, the prediction horizon should be either 1 or infinity to improve the control performance, depending on the relative difference between the modeling error and the terminal value function error.

Keywords: Model validation, Model reduction, Model predictive and optimization-based control
Abstract: Reactivity controlled compression ignition (RCCI) technology has gained in popularity due to its ability to achieve low level NOx and soot emissions with relatively high brake thermal efficiency. However, control of RCCI combustion is a complex task. Nonetheless, this challenge can be overcome with model-based control design (MBCD). In this work, a linear physics-based time-varying RCCI combustion model was developed and improved with an addition of a start-of-combustion (SOC) model. The model we developed, which is capable of real-time simulations, can predict the combustion phasing, heat-release, and cylinder pressure of an RCCI marine engine. The model showcases the trend-wise, high accuracy estimation of cumulative heat-release and cylinder pressure. Additionally, it is able to predict combustion phasing parameters with an error of less than 1% for control design.

Keywords: Dynamic networks, Machine learning, Time series modelling
Abstract: Round-trip time (RTT) and throughput are two of the most important parameters that are always monitored in networks with transmission control protocol (TCP). Deep learning-based time-series forecasting methods such as long-short term memory (LSTM) have recently been widely applied in TCP state prediction due to their strong pattern recognition and accurate prediction ability. However, the practical network environment can be dynamic and may deviate from the situations in which the deep model has been trained, resulting in deteriorated predictions. Furthermore, online retraining of the deep model to adapt to current working environment is usually unfeasible due to the nature of heavy computational complexity. In this paper, we propose a method which can online rectify the TCP predictions with a very small computational overhead (time consumption) by combining Bidirectional LSTM (BiLSTM) with the weighted regularized extreme learning machine (WRELM). Experiments show that the proposed method can greatly increase the online prediction accuracy of TCP states especially when the knowledge of the trained deep model diverges from the conditions of its original working environment.

Keywords: Time series modelling, Recursive identification, Randomized methods for modeling, identification and signal processing
Abstract: One of the challenges of cyber-physical systems is the data acquisition in real-time through different communication protocols within the scope of the Industrial Internet of Things (IIoT). This IIoT data is mainly specified as big in amount, continuous in nature, unstructured, and ambiguous format. It is also enigmatic in terms of structure for an overview of the manufacturing plant and which data belongs to which sensor or machine. Presently, relational database technologies are inadequate to cater to such kinds of IIoT data due to their strict data structure and relations. For this reason, we presented the ontology-based data repository model implementation for the process industry using MongoDB, which is a cross-platform document-oriented, most sought-after ample data storage database system nowadays. Firstly, we have designed a JSON-based data-model schema for the IIoT data repository, which gathers the data with the help of one of the most commonly used communication protocols within the scope of IIoT, OPC UA. Secondly, our schema stores the meta-data and data using embedded referencing through the same data-model. Hence, the scalability, flexibility, and variety of the data storage system have increased. Lastly, with the execution of insertion and search queries, we show how fast to get the desired data element hierarchy or track any particular data, which becomes possible due to the proposed ontology. This effectiveness of the ontology-based data-model also minimize the overhead of the machine learning engineer job.

Keywords: Time series modelling, Subspace methods
Abstract: In this paper, a novel compact dynamic inner canonical correlation analysis (DiCCA) algorithm with autoregressive integrated moving average (ARIMA) inner models is proposed to better capture the latent dynamics of high dimensional time series. It can extract latent factors that capture the underlying dynamics of the data and model them using the ARIMA structure, which has fewer parameters and more flexibility than the potentially high-order AR structure used in the original DiCCA algorithm. The proposed algorithm can also handle nonstationary latent factors by explicitly modeling the unit roots. The algorithm integrates the extraction and the modeling of a latent factor in one step, resulting in a consistent inner and outer model. The algorithm is applied to an industrial dataset and shows better performance than the original DiCCA algorithm.

Keywords: Fault detection and diagnosis, Time series modelling
Abstract: Deep learning (DL) has been known as one of the effective techniques for building data-driven fault detection methods. The successful DL-based methods require the condition that massive labeled data are available, but this is sometimes an inevitable obstacle in real industrial environments. As one of the solutions, autoencoders (AEs) are widely adopted since AEs can extract features from unlabeled data. However, some challenges in AE-based fault detection methods remain, such as the design of encoder architecture, the computational cost, and the usage of the limited labeled data. This paper proposes a new industrial fault detection method through learning instance-level representation of time-series based on the self-supervised contrastive learning framework (SSCL). The proposed method uses dilated-causal-convolution-based encoder-only architecture to extract the information from industrial time-series data. A new data augmentation method for time-series data is proposed based on the temporal distance distribution, which is used to construct positive pairs in SSCL. Moreover, the encoder is alternately trained by the new weighted contrastive loss and the traditional classification loss. Finally, the experiments are conducted on the industrial data set and a semi-physical system, showing the effectiveness of the proposed method.

Keywords: Grey box modelling, Machine learning, Nonlinear system identification
Abstract: By one of the most fundamental principles in physics, a dynamical system will exhibit those motions which extremise an action functional. This leads to the formation of the Euler-Lagrange equations, which serve as a model of how the system will behave in time. If the dynamics exhibit additional symmetries, then the motion fulfils additional conservation laws, such as conservation of energy (time invariance), momentum (translation invariance), or angular momentum (rotational invariance). To learn a system representation, one could learn the discrete Euler-Lagrange equations, or alternatively, learn the discrete Lagrangian function mathcal{L}_d which defines them. Based on ideas from Lie group theory, we introduce a framework to learn a discrete Lagrangian along with its symmetry group from discrete observations of motions and, therefore, identify conserved quantities. The learning process does not restrict the form of the Lagrangian, does not require velocity or momentum observations or predictions and incorporates a cost term which safeguards against unwanted solutions and against potential numerical issues in forward simulations. The learnt discrete quantities are related to their continuous analogues using variational backward error analysis and numerical results demonstrate the improvement such models can have both qualitatively and quantitatively even in the presence of noise.

Keywords: Time series modelling, Machine learning, Nonlinear system identification
Abstract: This paper investigates the performance of automatic sleep stage classification with automatically selected features from electroencephalographic (EEG) signals using a Convolutional Neural Network (CNN) based on 5- and 2-class models. We defined two ways for 2-class stratification, to classify the sleep stages and compare its performance with predictions made using a 5-class model. All the models were created using a CNN called EEGNet, and the experiments were carried out with the ISRUC-Sleep public dataset, which consist of 100 subjects and 6 EEG channels. With a single 5-class model for the entire dataset, we have obtained an average area under the receiver operating characteristic (AUROC) of 0.948. In the best case, we have obtained an average AUROC of 0.964, 0.967, 0.982 and 0.929 for the stratified 2-class models: Awake vs Sleep (Rapid Eye Movement (REM) + Non-REM [N1+N2+N3]), REM vs Non-REM (N1+N2+N3), (N1+N2) vs N3, and finally N1 vs N2. We have shown that in the four 2-class stratification-based models in a row, we can achieve an average AUROC of 0.97. The results obtained are promising and can lead to possible combinations of the 5- and 2-class models to improve the automatic sleep stage classification.

Keywords: Optimal control of partial differential equations, Control of heat and mass transfer systems
Abstract: We consider a Linear Quadratic Regulator (LQR) for the heat equation on the unit disk using distributed boundary control. This leads to a Riccati PDE for the kernel of optimal cost. In a simple case we are able to explicitly solve this PDE using Fourier series and compute the closed loop eigenvalues and eigenfunctions.

Keywords: Control of heat and mass transfer systems
Abstract: A material independent melt temperature controller for an extruder is presented. The extruder cylinder temperatures are varied based on the observed heat flow profile, which is estimated by a smart sensor, so that a desired melt temperature is obtained and the heat flow profile adopts a desired shape. The cylinder temperatures are controlled by a subordinate control concept. The concept is validated by extrusion experiments.

Keywords: Output regulation for distributed parameter systems, Control of heat and mass transfer systems, Stability of distributed parameter systems
Abstract: The paper considers robust output regulation of a Guyer-Krumhansl (GK) heat conduction law under modeling mismatches. More specifically, we consider the differences of the GK model compared with a simple diffusion equation (also known as the Fourier heat conduction law) and treat them as modeling mismatches in view of robust output regulation. We will consider a simple internal model based controller to solve the output regulation problem and analyze its robustness with respect to the differences between the two heat conduction laws. Moreover, we will test the robustness of the considered controller in numerical simulations.

Keywords: System identification and adaptive control of distributed parameter systems, Identifiability, Process modeling and identification
Abstract: The aim of this paper is to estimate the parameters of a counterflow heat exchanger, in the case where the dynamics are described by a system of two hyperbolic partial differential equations (PDEs). This study is performed using data collected on a laboratory test bench. First, we study the practical identifiability of the system, the objective of which is to verify the possibility of determining the parameters of the system in a unique way from the available data. Next, the parameter estimation is performed by a nonlinear optimization method, typically by the Levenberg-Marquardt algorithm on a set of ordinary differential equations obtained from the reduction of the original PDEs by the method of lines. Since the Levenberg-Marquardt algorithm may only converge with a reasonable estimate of the parameters, we first consider a least squares parameter estimation approach based on the global difference model. Finally, the results of this identification are validated by using laboratory data, and by verifying some statistical properties.

Keywords: Thermal and process control applications of distributed parameter systems, Nonlinear observers and filter design, Observer design
Abstract: Concrete is one of the most important building materials and is used as a collective term for different dispersions of cement and aggregates. High performance concrete (HPC) is a rather new type of concrete mixture. A topic of current research is the behavior of fresh concrete during heat treatment. It is known that the heat treatment of fresh concrete accelerates the reaction of the cement and thus the chemically complex process of hydration. The mathematical representation of the curing concrete is a further step towards the systematic investigation. The objective of this work is the development of a state estimator to determine the temperature behavior of the fresh HPC mixture under heat treatment. Using a continuum representation describing the spatial-temporal evolution of temperature, moisture, and maturity of fresh HPC during heat treatment a finite-dimensional approximation is derived using a high order finite difference scheme. Experimental data is included for model parameterization by optimization. Based on this, three different nonlinear extensions of the Kalman filter (KF) techniques are realized and compared combining simulated and experimental data.

Keywords: Predictive control, Linear multivariable systems, Process control
Abstract: This paper proposes an offset-free distributed implementation of a model predictive controller that employs fuzzy negotiation between agents. The scheme is based on model augmentation with additional disturbances to enable zero-offset tracking. Moreover, we code the negotiation criteria as a set of suitable fuzzy rules and consider stability and feasibility guarantees in the controller design for the linearized subsystems. We applied the method to an experimental four-tank plant, showing its effectiveness despite the coupling between subsystems and system-model mismatch.

Keywords: Regulation (linear case), Process control
Abstract: This paper treats the problem of feedforward control from measurable load disturbances. Most of the tuning rules based on IAE or ISE optimization for feedforward compensators that have been presented in recent years are compared with each other as well as with optimal compensators. There are two different feedforward control schemes that may be used, and these schemes are also compared and analyzed. The paper gives recommendations for selection of both control schemes and tuning rules.

Keywords: Linear systems, Time-delay systems
Abstract: In this paper we propose a delay-dependent analysis of the positive invariance property with respect to linear discrete-time systems with delayed states. An appropriate model transformation is employed, together with a matrix parametrization, which allow the derivation of delay-dependent invariance conditions of polyhedral sets with respect to the transformed model. We then show that such conditions imply the confinement of the state trajectories of the original system in the set, as long as the initial states satisfy additional constraints related to the system dynamics. The characterization of this set of admissible initial conditions gives rise to the proposition of a less conservative definition of set-invariance. We illustrate through numerical examples the fact that, under the proposed definition, confinement of state trajectories in the set can be achieved even though it is not invariant according to the classical definition.

Keywords: Linear multivariable systems, Time-invariant systems, Robustness analysis
Abstract: In contrast to the well-developed gain analysis for multi-input-multi-output (MIMO) linear timeinvariant (LTI) systems, the research on the phase analysis does not share the same status. In this paper, we introduce the phase response of a class of discrete-time (DT) LTI multivariable systems by exploiting a definition of matrix phases based on the numerical range. Positive frequency joint sectorial systems are also defined, which can generalize the positive real and negative imaginary systems. The interpretation of system phases is given in terms of the input and output signals of the system. A sectored real lemma is obtained to characterize the phase information from a state–space realization. Motivated by finding a phasic counterpart to the small gain theorem, we develop a small phase theorem for the internal stability of a closed-loop system. The phase properties are also investigated for parallel and feedback interconnected systems.

Keywords: Linear systems, Systems with saturation
Abstract: This paper considers the problem of controlling a linear system affected by asymmetrical input saturation. The proposed solution is based on using a linear matrix inequality (LMI)-based methodology to find the gains of a switching state-feedback controller. The main difference and contribution when compared to existing approaches is that the switching rule is chosen based on the closed-loop performance that each of the non-saturating controller gains can achieve when used with the current value of the state vector. Although the main focus of the paper is on time-invariant systems, the possible extension to linear parameter-varying (LPV) systems is discussed. An illustrative example is used to show the main features of the proposed approach.

Keywords: Linear systems, Optimal control theory, Real-time optimal control
Abstract: Approaches to trajectory planning for lumped parameter models of flexible systems with kinematic input and state constraints are compared. These comprise so-called impulse-shaping techniques, a flatness-based approach, and optimal-control methods. The latter includes a novel real-time capable technique for almost time-optimal control. On the basis of a simple example, the different methods are compared in simulation studies w.r.t. the time required for typical rest-to-rest transitions, and, moreover, w.r.t. their robustness to parameter variations.

Keywords: Linear systems, Parametric optimization, Structural properties
Abstract: The paper proposes a tunable inerter and investigates its merits in suppressing building vibrations. We considered a five-layer building model and derived the optimal inertances for suppressing six historical earthquakes. We also constructed a testing rig for experimental verification. The results showed that the building could effectively repress different earthquakes by particular inerter settings. The experiments also confirmed the benefits of the tunable inerter for earthquake suppression of buildings.

Keywords: Fractional systems, Time-delay systems, Time-varying systems
Abstract: The robust fractional order oil flow control in an oil pipeline is addressed in this paper. An experimental identification based on a step like input was developed which yielded a second order transfer function with a zero and a time varying time delay. Controllers are designed next that have as specifications a nominal phase margin, zero steady state error in the step response and a robust damping, expressed in terms of a maximum allowed variation of the phase margin of ±10% around the nominal value. We have considered integer and fractional order controllers. Rules to tune the parameters of these controllers have been obtained and a condition to get minimum phase closed loop systems has been derived. We have shown that PI controllers verifying the three design conditions often yield undesired nonminimum phase closed loop systems, which is avoided using the proposed fractional order controllers. Simulations carried out in this paper illustrate these results.

Keywords: Fractional systems, Process control, Decoupling problems
Abstract: In this paper we investigate the performance that can be achieved by using Fractional-Order Proportional-Integral-Derivative (FOPID) controllers for Multiple-Input-Multiple-Output (MIMO) industrial processes. In particular, we consider several well-known interacting process models taken from the literature and we tune the controllers by minimizing an optimization functional via genetic algorithms. We consider two different cost functions, one based on the integrated absolute error and the other one on the integrated square error. A comparison with standard integer-order Proportional-Integral-Derivative (PID) controller is performed so that the advantages of using decentralized FOPID controllers for set-point tracking and decoupling can be clearly evaluated.

Keywords: Fractional systems
Abstract: This paper focuses on the identification of commensurate fractional order systems from binary measurements on the output. The proposed algorithm allows the estimation of both the parameters of the model and the commensurate order. The algorithm is based on a specific design of the output signal as a sum of sinusoids. It allows the parametrization of the output signal on a basis of sinusoids and the estimation of the high resolution output. Numerical simulation results are provided to show the efficiency of the method.

Keywords: Fractional systems, Observer design
Abstract: In this paper, a method for disturbance estimation for a class of non homogeneous time fractional advection dispersion equation with general boundary conditions, is proposed. The disturbances are considered to a ect the boundary conditions or the measurements or both. The proposed method is based on modulating function. Numerical simulations are presented to illustrate the performance of the approach.

Keywords: Fractional systems, Linear systems, Complex systems
Abstract: Modeling the global climate system through the temperature output is a very challenging task. Knowing the fractional operator long memory property to well model diffusion phenomena with very few parameters, it is proposed in this paper to use fractional models for climate change modeling. System identification is applied for continuous-time system identification of multiple-input single-output (MISO) fractional order models. When differentiation orders are assumed to be known, coefficients are estimated using the simplified refined instrumental variable method for continuous-time fractional models extended to the MISO case. For unknown differentiation orders, a two-stage optimization algorithm is proposed with the developed instrumental variable for coefficient estimation and a gradient-based algorithm for differentiation order estimation. Finally, the estimation with fractional models is carried out on real input/output data and provide a very good goodness fit of the global earth temperature.

Keywords: Trajectory and path planning, Multi-vehicle systems, Autonomous vehicles
Abstract: We present a novel method to address the problem of multi-vehicle conflict resolution in highly constrained spaces. An optimal control problem is formulated to incorporate nonlinear, non-holonomic vehicle dynamics and exact collision avoidance constraints. A solution to the problem can be obtained by first learning configuration strategies with reinforcement learning (RL) in a simplified discrete environment, and then using these strategies to generate new constraints and initial guesses for the original problem. Simulation results show that our method can explore efficient actions to resolve conflicts in confined space and generate dexterous maneuvers that are both collision-free and kinematically feasible.

Keywords: Decentralized control and systems, Human operator support, Autonomous robotic systems
Abstract: This paper presents a novel implementation of a human-swarm interface that allows humans to define an area with desired shape to be reached by a multi-robot system. Human-swarm interaction can be useful in order to exploit human intelligence and knowledge for the operation of swarm robots. The proposed work deals with limitations usually met when dealing with real-world implementation, e.g. limited sensing capabilities of the agents and hard conditions where communication is difficult or even completely denied. Gaussian Mixture Models are exploited in order to define an appropriate probability density function of the environment based on the area selected by a human operator. Then, velocity input for each robot is calculated in a distributed manner using Voronoi tessellation and Lloyd’s algorithm. Finally, results of both virtual and real-world tests are presented, showing the final configuration reached by the multi-robot system in comparison with the desired region defined on the graphical interface.

Keywords: Multi-vehicle systems, Vehicle dynamic systems, Nonlinear and optimal automotive control
Abstract: In this paper an unmanned aerial vehicle works cooperatively with a unmanned ground vehicle to aid in the automatic control of the ground vehicle. One challenge is when the aerial vehicle has to urgently rendezvous with the ground vehicle. An optimal control problem is formulated for such a case the key contribution is the singular optimal control automatic control formulation Illustrative results are utilised to demonstrated for this key formulation of singular optimal control solution.

Keywords: Multi-vehicle systems, Mission planning and decision making, Autonomous vehicles
Abstract: Searching and tracking mobile targets rely most often on modeling the uncertainty and various perturbations by stochastic processes. The detection and location of the targets are performed with Bayesian estimation, which reliability and resulting performance are deeply linked to the adequacy of the stochastic models. An alternative approach limits the representation of these perturbations by defining the bounds within which they can vary. Set-membership estimation techniques have been developed to handle this representation. This paper compares the performance of set-membership and stochastic Bayesian estimation techniques for target search and tracking for scenarios integrating false alarms. For this purpose, estimation schemes are presented for each approach. The ability of estimators to find real targets and not to be deceived by false targets or imperfect sensors are compared in simulations.

Keywords: Multi-vehicle systems, Autonomous vehicles, Decentralized control and systems
Abstract: This paper proposes a controller for safe lane change manoeuvres of autonomous vehicles using high-order control barrier and Lyapunov functions. The inputs are calculated using a quadratic program (CLF-CBF-QP) which admits short calculation times. The controller allows for adaptive cruise control, lane following, lane switching and ensures collision avoidance at all times. The novelty of the controller is the decentralized approach to the coordination of vehicles without switching of controllers. In particular, vehicles indicate their manoeuvres which influences their own safe region and that of neighboring vehicles. This is achieved by introducing so-called coordination functions in the design of control barrier functions. In a relevant simulation example, the controller is validated and its effectiveness is demonstrated.

Keywords: Decentralized control and systems, Human and vehicle interaction, Intelligent Transportation Systems
Abstract: Connected vehicles are expected to play a pivotal role in the future of road transportation. One of the major advantages of connected vehicles is their ability to perform cooperative movement, such as platooning, in order to ensure better safety and efficiency. In the near future, connected vehicles will have to co-exist and interoperate with other human-driven vehicles. This necessitates the study of cooperative movement of connected vehicles subject to the behavior of human-driven vehicles. We study a platooning scenario with two different vehicle types: (i) connected cars (CCs) and (ii) human-driven motorized two-wheelers (MTWs). The CCs are automated and can share information with each other, while a human-driven MTW does not have any level of automation. An MTW may either follow the CC ahead, i.e. participate in the platoon, or choose to swerve away. The objective of this paper is to design a decentralized controller for the CCs that can ensure a safe headway with respect to the preceding vehicle, even under the swerving behavior of motorcyclists. The interaction between the CCs and behavior of the MTWs are captured using a random graph. The Algebraic Riccati equation is then used to derive a linear consensus controller that achieves the desired platoon spacing. The proposed design is corroborated through numerical simulations.

Keywords: Autonomous robotic systems, Information and sensor fusion, Guidance navigation and control
Abstract: This paper examines the generation of visibility maps for the purposes of robotic path planning. The paper proposes a map construction technique that extends a version of the viewshed problem to calculate the rays to and from all discrete points on the map efficiently. The introduced form is parallelizable, making it suitable for fast GPU computation necessary for real-time calculation during path planning. New algorithms are introduced to generate worst- and best-case visibility maps by bounding precision uncertainties in the discretized map. Visibility theory is then applied to quantify the detectability of each viewpoint in the space based on atmospheric visibility models. Path planning is executed on the generated cost map illustrating its usability for visibility path planning.

Keywords: Autonomous robotic systems
Abstract: A reinforcement learning (RL) enabled intelligent motion planning for collision-free autonomous docking manoeuvre explicitly designed for a robotic floating satellite emulation platform is presented in this article. The Twin Delayed Deep Deterministic Policy Gradient-based RL algorithm involving deep neural network architecture in the actor-critic framework is considered to obtain the collision-free safe docking policy. The RL agents have been trained to perform a resilient target acquisition, ensuring its terminal position and velocity requirements while enabling the capability to avoid both static and dynamic obstacles. The resulting optimal policy is implemented as a feedback control law to enable computationally efficient onboard reactive motion planning for autonomous safe docking of the robotic floating satellite platform in a complex dense debris environment. The efficacy of the proposed motion planning scheme is validated with numerous simulation studies, where it is depicted that the trained RL-based planner has the potential to address the target acquisition with a sufficient degree of accuracy in the presence of both static and dynamic obstacle scenarios.

Keywords: Autonomous robotic systems, Nonlinear adaptive control, Constrained control
Abstract: In this article, a novel adaptive controller is designed for Euler-Lagrangian systems under predefined time-varying state constraints. The proposed controller could achieve this objective without a priori knowledge of system parameters and, crucially, of state-dependent uncertainties. The closed-loop stability is verified using the Lyapunov method, while the overall efficacy of the proposed scheme is verified using a simulated robotic arm compared to the state of the art.

Keywords: Autonomous robotic systems, Mechatronics for mobility systems, Intelligent robotics
Abstract: This paper proposes a principle of fully autonomous ground mobile landing recovery of Unmanned Aerial Vehicles (UAV) for the problems of relatively fixed landing point, passive recovery, poor flexibility, and environmental adaptability, which mainly includes localization, landing point tracking, and buffering landing for quadrotor UAV. Firstly, aiming at the problem that it is difficult to accurately obtain the position of a UAV in dynamic mobile landing recovery, a target location method based on Asynchronous Multisensor Information Fusion(AMIF) and servo turntable focus tracking is proposed. Secondly, to achieve fast and high-precision tracking of UAVs, a tracking control strategy of an independently driven landing recovery system and a Stewart six-degree of freedom platform is proposed. Then, to solve the problems of large impact force and center of gravity instability in the landing process of UAV, a stationarity control algorithm based on model prediction and a compliance control algorithm based on adaptive variable impedance are designed to achieve active compliance control while adjusting the position and attitude of the receiving surface in real-time. Finally, a quadrotor unmanned landing and recovery experimental platform is built to verify the feasibility of the ground mobile landing and recovery strategy proposed in this paper and the effectiveness of the control algorithm.

Keywords: Autonomous robotic systems, Robots manipulators, Field robotics
Abstract: Cooperative robotic systems are becoming increasingly important, as cooperation between two or more robots makes it possible to tackle tasks that are difficult, if not impossible, to perform with a single robot. Among the various cooperative robotic systems, in this paper we focus on a heterogeneous robotic system to perform a building activity where large and heavy blocks are used. The proposed robotic system is based on a lifting mechanism and a robotic arm. This system exploits the characteristics of the lifting cable mechanism to hold most of the weight of the block, while the use of the rigid robot allows to obtain the desired precision during the fine placement. In this paper, we first derive the dynamic model of the system under consideration. Then, the kinematics analysis of the multi-robot system is investigated. Based on these two previous results, a control scheme is designed to ensure safe and correct cooperation between the two robotic systems. Experimental results with an in-scale prototype of the proposed robotic solution show the efficiency of our approach.

Keywords: Robots manipulators, Autonomous robotic systems
Abstract: Hand-eye calibration is a critically important process for vision-based control of robot manipulators. It involves obtaining transformation matrices between the camera and either the robot base or the end-effector. The aims of this research are to develop algorithms capable of performing automated hand-eye calibration with 2D cameras. An automated sequential process for the hand-eye calibration is proposed by employing Kalman filtering in recursive computation. In particular, the proposed method has been implemented in eye-in-hand and eye-to-hand systems. The automated algorithm developed in this research has been effectively verified in experiments. Based on experimental results, the proposed algorithm results in translation and rotation errors less than 1.5 mm and 0.7 degree after 25 iterations, respectively.

Keywords: Mechatronic systems, System analysis and optimization, Motion control systems
Abstract: This paper presents the mechatronic system design and control of a tip/tilt motion platform with a magnetically levitated mover tailored for dynamic rotational scanning operation. It employs a Lorentz force-based actuation system with five degrees of freedom and four spherical Halbach arrays on a fully passive mover. The system includes highly integrated optical proximity sensors for sensing the mover position in all degrees of freedom and a passive magnetic bearing for restraining the non-actuated degree of freedom. The actuation system enables a mechanical angular range of +/-1° in both axes, with first structural modes of the 3D printed mover occurring at 1 kHz. A control system is designed for the five actively controlled degrees of freedom for obtaining a robustly stable and precise motion system. The resulting closed-loop system has a bandwidth of 200 Hz in the rotational and 100 Hz in the translational degrees of freedom and a small positioning uncertainty of 0.02% of the full motion range.

Keywords: Mechatronic systems, Embedded robotics
Abstract: Fuzzy Logic (FL) is well-known as an intuitive framework to tackle imprecision and uncertainty while allowing to model expert's knowledge in rule-based control. The objective of this study is to investigate the design of embedded fuzzy control systems combining fast execution and parallel processing capabilities provided by Field Programmable Gate Arrays (FPGAs) and Reconfigurable Inputs/Outputs (RIO) boards. We implemented a suitable fixed-point FL to realize a computationally efficient control framework on an FPGA target. As such, the paper provides a concise method to deploy Fuzzy Logic Control (FLC) using RIO-FPGA technology. We also provide a technique for implementing the three stages that constitute the FLC structures in LabVIEW environment using fixed math operations. The experimental work involves tracking the trajectory of a mobile robot and shows the feasibility and the efficiency of the proposed FLC-FPGA controller in comparison to the traditional PID-FPGA controller, along with the added benefits of fuzzy control frameworks.

Keywords: Mechatronic systems, Vibration control, Mechatronics
Abstract: This paper proposes a sample-tracking vibration isolator that integrates a flexure-guided hybrid reluctance actuator (HRA) with a negative stiffness for quasi-zero stiffness (QZS). In a sample-tracking vibration isolator, for example, for inline metrology, an actuator carries and positions an instrument with respect to a sample. By maintaining the distance between the instrument and the sample by feedback control, the vibrations between them are rejected for measurements in vibrational environments. To achieve high vibration isolation and rejection, the proposed sample-tracking vibration isolator uses the negative stiffness that is magnetically realized and cancels a flexure stiffness for QZS. Due to the rigid magnetic components of the HRA such as a permanent magnet and ferromagnetic cores, a high second mechanical frequency of 2.1 kHz is realized. Consequently, a high crossed-loop control bandwidth of 806 Hz is achieved with feedback control. In order to validate the effectiveness of the rigid negative stiffness of the HRA, the negative stiffness is tuned to cancel a flexure stiffness at experiments. Experimental results show that transmissibility at a low frequency of 10 Hz is significantly decreased from -34 dB to -60 dB by a factor of 20 and successfully demonstrate high vibration isolation performance of the sample-tracking vibration isolator using the negative stiffness of the HRA.

Keywords: Mechatronics, Mechatronic systems, Motion control systems
Abstract: The magnetic suspension of a ferromagnetic object (floator) across a conductive wall was achieved without directly sensing the displacement of the floator. When a floator is inside a chamber made of conductive material, the cross-wall detection of the motion of the floator is impossible or at least difficult with an eddy-current type displacement sensor. This study discussed two approaches to the problem. In the first approach, an electromagnet directly actuates the floator across the wall. In the second approach, double series magnetic suspension was applied. Because a critical problem due to induced voltage is expected in the former, the latter was mainly treated. In the latter system, the motion of the first floator outside the chamber is detected directly. In contrast, the motion of the second floator inside the chamber is not detected. Instead, the displacement and velocity of the second floator is estimated by an observer, which is referred to as indirect sensing. Because the system is controllable and observable even with indirect sensing, the system can be stabilized by feeding back these estimated signals. To demonstrate the validity of such an approach, an experimental apparatus was designed and manufactured. A full-order observer was built to estimate the displacement and velocity of the second floator. The displacement estimated by the observer agreed with the actual displacement in some degree. Then, magnetic suspension with indirect sensing was achieved by replacing the sensor signal by the estimated signal.

Keywords: Mechatronic systems, Bayesian methods, Sensors and actuators
Abstract: Sensors in high-precision mechatronic systems require accurate calibration, which is achieved using test beds that, in turn, require even more accurate calibration. The aim of this paper is to develop a cascaded calibration method for position sensors of mechatronic systems while taking into account the variance of the calibration model of the test bed. The developed calibration method employs Gaussian Process regression to obtain a model of the position-dependent sensor inaccuracies by combining prior knowledge of the sensor with data using Bayesian inference. Monte Carlo simulations show that the developed calibration approach leads to significantly higher calibration accuracy when compared to alternative regression techniques, especially when the number of available calibration points is limited. The results indicate that more accurate calibration of position sensors is possible with fewer resources.

Keywords: Mechatronic systems, Output feedback control (linear case), Modeling
Abstract: Dynamic error budgets are an essential tool in identifying opportunities for improvements in a control system for Gravitational Wave detectors, but their potential is often not fully utilized in the control design. This paper presents a model and dynamic error budget for a challenging nested control system in the Advanced Virgo detector in combination with a systematic control design framework for one of the controllers. This framework fully utilizes the dynamic error budget by using H2 synthesis to allow for fast iterations in the control design when dealing with conflicting control objectives. Simulations together with experimental results on Advanced Virgo illustrate the effectiveness of the presented framework.

Keywords: Statistical analysis, Machine learning, Estimation theory
Abstract: Recent works empirically demonstrated that incorporating target derivatives, in addition to the conventional usage of target values, during the training process improves the accuracy of the predictor and data efficiency. Despite the successful application of gradient data in the learning process, very little is understood theoretically about their performance guarantee. In this paper, our goal is to highlight (i) the limitations of gradient data on their performance guarantees, especially in low-data regimes, and (ii) the extent to which the gradients affect the learning rate. Our result implies that in a low-data regime, if the Lipschitz of the target function is below a threshold, gradient data for Sobolev training outperforms the classical training in terms of sample efficiency. For a target function with a large Lipschitz constant, there is a threshold for training data size beyond which the gradient data perform better than conventional training. The convergence behavior of gradient data for Sobolev training is studied, and the learning rate of order mathcal{O}(n^{-frac{1}{2}+epsilon}) is derived. Experiments are conducted to determine the effect of gradient data in the learning process.

Keywords: Stochastic control, Realization theory
Abstract: This paper considers the problem of steering an arbitrary initial probability density function to an arbitrary terminal one, where the system dynamics is governed by a first-order linear stochastic difference equation. It is a generalization of the conventional stochastic control problem where the uncertainty of the system state is usually characterized by a Gaussian distribution. We propose to use the power moments to turn the infinite-dimensional problem into a finite-dimensional one and to present an empirical control scheme. By the designed control law, the moment sequence of the controls at each time step is positive, which ensures the existence of the control for the moment system. We then realize the control at each time step as a function in analytic form by a convex optimization scheme, for which the existence and uniqueness of the solution have been proved in our previous paper. Two numerical examples are given to validate our proposed algorithm.

Keywords: Stochastic control, Synthesis of stochastic systems, Randomized methods for stochastic systems
Abstract: This paper addresses the problem of steering the distribution of the state of a discrete-time linear system to a given target distribution while minimizing an entropy-regularized cost functional. In our previous work, we dealt with the case where the running cost contains only a control cost and derived the explicit form of the optimal policy. Its analysis is based on two coupled Lyapunov equations. On the other hand, in the present work, we consider the distribution control problem where a quadratic state cost is also present. Then, we reveal that our problem boils down to solving coupled Riccati equations instead of the Lyapunov equations. In addition, we give the explicit form of a solution of the coupled Riccati equations.

Keywords: Stochastic control, Decentralized control and large-scale systems, Optimal control theory
Abstract: We investigate a class of mean field games containing a large number of major and minor players. Each player minimizes a quadratic-tracking type risk-sensitive cost functional, where the reference signal is a function of the state average term of the major and minor players. To reduce the complexity for solving the problem, we design a sequence of decentralized strategies by the Nash certainty equivalence principle. Firstly, we apply the two-layer state aggregation method to construct the fixed-point equations for the estimations of the state average terms and give the conditions for the existence and uniqueness of the fixed points. Then, we design a sequence of decentralized strategies by the estimations of the state average terms based on local information. It is shown that the estimations of the state average terms are consistent with the true values for the closed-loop systems, and the sequence of strategies designed is a decentralized asymptotic Nash equilibrium.

Keywords: Stochastic control, Consensus and reinforcement learning control, Multi-agent systems
Abstract: We study Nash equilibria learning of a general-sum stochastic game with an unknown transition probability density function. Agents take actions at the current environment state and their joint action influences the transition of the environment state and their immediate rewards. Each agent only observes the environment state and its own immediate reward and is unknown about the actions or immediate rewards of others. We introduce the concept of weighted asymptotic Nash equilibrium with probability 1 and design a two-loop algorithm by the equivalence of Nash equilibrium and variational inequality problems. In the outer loop, we sequentially update a constructed strongly monotone variational inequality by updating a proximal parameter while employing a single-call extra-gradient algorithm in the inner loop for solving the constructed variational inequality. We show that if the associated Minty variational inequality has a solution, then the designed algorithm converges to the k^{1/2}-weighted asymptotic Nash equilibrium.