Keywords: Cognitive aspects of automation systems and humans
Abstract: The question is essential and not just rhetorical. It is essential because we tend tacitly to accept the statement of William Thompson (aka Lord Kelvin) that:”Qualitative knowledge is real, but... quantitative knowledge is almost always better.” Since 1891, We have therefore conventionally but wrongly assumed that we cannot understand something unless we can measure it, even though it actually is the other way around. Namely, that we cannot measure something unless we fully understand it first. Quality must therefore precede quantity. Whether resilience is one or the other also leads to different questions, if resilience is assumed to be a quantity, the essential question is simply “how much resilience is there”. “But if resilience is a quality the essential question becomes just “how does resilience come about”. The answer to the latter question is clearly more important for engineering, for design and for control!

Keywords: Information and sensor fusion, Intelligent robotics, Mobile robots
Abstract: Volume estimation in large indoor spaces is an important challenge in robotic inspection of industrial warehouses. We propose an approach for volume estimation for autonomous systems using visual features for indoor localization and surface reconstruction from 2D-LiDAR measurements. A Gaussian Process-based model incorporates information collected from measurements given statistical prior information about the terrain, from which the volume estimate is computed. Our algorithm finds feasible trajectories which minimize the uncertainty of the volume estimate. We show results in simulation for the surface reconstruction and volume estimate of topographic data.

Keywords: Intelligent robotics, Autonomous robotic systems, Robotics technology
Abstract: Tactile robots can perform complex interaction skills, e.g., polishing. Such robots should therefore be designed to be adaptive to environmental uncertainties such as changing geometry and contact-loss. To address this, we propose a tactile exploration technique to observe the local curvatures of the physical constraints such as corners, edges, etc. for updating predefined tactile skill policies accordingly. First, we develop a unified force-impedance control approach in which the force controller significantly improves the geometry following performance due to the ensured contact. Second, we use the proposed controller to autonomously investigate the unknown environment via the local curvature observer, designed to be a dynamic process. Finally, the exploration performance of the proposed controller is demonstrated by using a polishing skill on an unknown 3D surface, where the robot is observed to autonomously investigate the unknown surface from top to bottom along the edges and corners.

Keywords: Intelligent robotics, Perception and sensing
Abstract: The automation of waste cranes has been demanded to perform garbage incineration work with fewer workers efficiently. In particular, data-driven learning approaches are desirable for waste crane automation. When deciding whether to re-grab garbage by a waste crane to grab more garbage, human operators are efficiently making decisions based on visual information. However, the current automation system has to lift up the bucket to measure grabbed garbage weight to decide re-grab. The lifting motion after grabbing makes the crane motion inefficient. For this limitation, we propose a re-grabbing decision system with feedback from in-bucket camera images for the efficiency of waste crane automation by introducing the vision sensor in the bucket. To simplify the decision process of re-grabbing, we separate the re-grabbing decision system from the in-bucket image into grabbed garbage weight estimation and the re-grabbing decision policy based on estimated garbage weight. Moreover, the weight estimator model and the re-grabbing policy model are designed based on a Bayesian manner for data efficiency. The effectiveness is verified in a robotic waste crane system with an in-bucket camera. We confirmed the proposed method could learn re-grabbing decisions from the autonomously collected data. It achieved more efficient re-grabbing by waste cranes than the conventional automated system.

Keywords: Mobile robots
Abstract: This paper presents an inertial sensor-based localization method using a deep neural network (DNN)-based velocity estimator. Among various sensors commonly used in localization, inertial sensors are less affected by the surrounding environment. The inertial sensor consists of an accelerometer and a gyroscope, which can estimate the poses of a robot. However, inertial sensors need to be used with other sensors for localization because inherent large drift errors are difficult to prevent. To overcome this problem, a DNN-based velocity estimator that reduces the position error by learning the data pattern of the inertial sensor and by limiting the range of the estimated velocity is proposed. The DNN-based velocity estimator comprises a convolutional neural network, a fully connected layer, and a smoothing filter. To limit the range of the estimated velocity, the range of velocity is divided by the total number of classes, and the divided range is assigned to each class. The relationship between consecutive classes is learned using a smoothing filter. Dataset-based experiments are performed to train the DNN-based velocity estimator and to evaluate the performance of the proposed localization. The experimental result shows that the proposed localization reduces the position errors by 99% compared to integration-based localization and by 91% compared to extended Kalman filter-based localization.

Keywords: Autonomous robotic systems, Perception and sensing, Robotics technology
Abstract: Simultaneous Localization and Mapping (SLAM) is a fundamental technology for robotics. Vision-based SLAM has been developed for many years, but it is still difficult for SLAM system to handle dynamic environments. In this paper, we present GMP-SLAM, a real-time RGB-D SLAM system for highly dynamic environments with the help of GPU Grid Map Projection— a GPU dynamic points detection method we design. SLAM is a time-sensitive system for robotics, and it is hard to reach real time, especially in dynamic environments because it is necessary to track moving objects and it takes a lot of time. GMP-SLAM is based on ORB-SLAM2, which is one of the best feature-based SLAM frameworks and can reach real time just in CPU. But ORB-SLAM2 cannot handle highly dynamic environments very well, and most work focus on tracking moving objects with a neural network, which cannot reach real time even with the help of GPU. To solve real-time problem, we propose an all-in-parallel dynamic points detection framework for visual simultaneous localization and mapping (VSLAM) in dynamic environments based on 3D occupancy grid maps. Our SLAM system can provide not only higher trajectory accuracy but also a 3D grid map for navigation. We test our SLAM system in our real-world datasets we record and get higher trajectory accuracy than ORB-SLAM2. At the same time, our system can run nearly in 20Hz, which is much better than existing VSLAM framework in dynamic environments.

Keywords: Intelligent robotics, Mobile robots, Evolutionary algorithms in control and identification
Abstract: This work proposes to use a teach-and-repeat method to navigate in an outdoor environment. Instead of using state-of-the-art feature point descriptors to match the current images to the database, we rely on the Generated Binary Robust Independent Elementary Features (GRIEF), a model robust to changes in illumination and weather. These descriptors are obtained via a training process employing evolutionary methods and binary comparison tests. The objective is to obtain visual feature descriptors designed for a given task. For the teach-and-repeat task, they are trained to optimize the estimate of the robot heading error with respect to a learned path. In this work, we propose to investigate how Differential Evolution (DE) can improve the performance of the GRIEF in terms of the convergence rate during the training phase and of fitness score. The new approach, named GRIEF-DE, is trained with the Michigan data set, made of pictures of an outdoor environment with significant changes in appearance caused by seasonal weather variations and illumination, and compared with the original GRIEF. The proposed model was applied to 4 different data sets and the experimental results suggest that the use of Differential Evolution improves the training performance as well as the estimation of the robot orientation with respect to the path of reference.

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

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

Keywords:
Abstract: Shotgun presentations from Manufacturing Plant Modelling and Control I (Interactive Session). Each talk will be for a 2 minutes duration.

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

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

Keywords:
Abstract: Shotgun presentations from Manufacturing Plant Modelling and Control II (Interactive Session). Each talk will be for a 2 minutes duration.

Keywords: Lyapunov methods, Stability of nonlinear systems, Control of interconnected systems
Abstract: This paper aims to develop a dissipativity-based framework for analysis of interconnection of scalar systems defined for positive variables exhibiting interior equilibria. To analyze asymptotic stability of an interconnected system, dissipativity-based approaches aggregate supply rates of component systems. This paper focuses on the fact that changes of parameters in component systems having interior equilibria can violate not only the stability, but also the positivity of variables. This paper peruses supply rates whose combination leads to a criterion with which the stability and the positivity can be verified simultaneously. Properties of supply rates suitable for logarithmic storage functions are investigated. It is explained how stability analysis leads us to a wrong conclusion if the positivity is forgotten. Based on those observed essentials, this paper proposes several asymmetric supply rates carrying the positivity information. Their usefulness is demonstrated through global and semi-global analysis of feedback interconnections, cyclic networks and a model of infectious diseases.

Keywords: Positive systems, Linear systems, Time-varying systems
Abstract: This paper extends the theory of positive systems to sampled-data systems. When considering positivity of LTI systems, where to take the initial time is unimportant. However, since sampled-data systems are periodically time-varying systems from the perspective of the continuous-time input and output signals, it is important to consider which time should be viewed as the initial time. Therefore, this paper studies positivity of sampled-data systems first by considering the case where a sampling instant is viewed as the initial time, and secondly by further considering the case where an arbitrary intersample instant is viewed as the initial time. The two different situations lead to quite relevant issues of what to take as the initial conditions and what to assume on the initial conditions in appropriately defining the internal/external positivity of sampled-data systems. These arguments are combined to derive the necessary and sufficient conditions for ultimate notions of (i.e., initial-time-independent) internal/external positivity of sampled-data systems.

Keywords: Positive systems, Linear systems
Abstract: A linear system is said to be externally positive if the system output is non-negative for all time when driven by a non-negative input from the zero initial state. External positivity is well known to be equivalent to non-negativity of the impulse response and hence monotone nondecreasing step response. Despite the apparent simplicity of characterizing systems with non-negative impulse response, the determination of necessary and sufficient conditions for external positivity from a given transfer function is a long-standing open problem. In this paper, we propose a method which approximately characterizes the true region capturing the numerator polynomials of all (strictly proper) externally positive linear systems whose specified poles are assumed to satisfy a known necessary condition for external positivity. We compute an (outer) approximation of the true region via the construction of a convex polytope, each facet of which is contained in a supporting hyperplane of the true region. The proposed method requires only modest computational effort; has an accuracy which can be increased readily; applies to systems having orders n geq 4 for which no general characterizations of external positivity are currently known; and handles systems with complex poles (possibly repeated). Numerical examples illustrate the effectiveness of the proposed method.

Keywords: Positive systems, Observers for linear systems, Linear multivariable systems
Abstract: In this paper are solved problems concerning the state estimation in observer-based control of ostensible positive linear systems. The method is based on a composed form of the system matrix representation, where the constructive procedure is given to the observer synthesis. The method is flexible and allows to obtain the observer with the gain matrix being strictly positive. The design is computationally simple and reduces to feasibility of the LMI problem. Numerical simulation is presented to illustrate the validity of the proposed method.

Keywords: Positive systems, Linear systems, Optimization and control of large-scale network systems
Abstract: This paper is concerned with the optimization of the impulse-to-peak value of positive linear systems. Although the results of the analysis of the peak value of positive linear systems have been reported in the literature, the synthesis problem of impulse-to-peak value remains unresolved. To address this problem, in this paper we show that a cost-constrained minimization problem of the impulse-to-peak value of parameterized positive linear systems can be reduced to a DC (difference of convex functions) program. The derivation utilizes the log-log convexity of posynomial functions and a recent characterization of the impulse-to-peak value of positive systems via a linear program. To illustrate the effectiveness of our results, we study the optimization problem for minimizing the peak infection of networked epidemic spreading processes.

Keywords: Healthcare management, disease control, critical care, Control in system biology, Systems biology
Abstract: How to maximise socioeconomic activities while retaining the effect of COVID-19 endemic under control is an essential topic when many countries have decided to live with COVID. In this paper, we show how optimal control can be used to classify the importance of non-pharmaceutical interventions (NPIs) to counter an epidemic under different expectations. We focus on the modelling of small environments, such as schools, universities, and care homes, where the population can be separated into two distinctive subgroups. We propose a model for such a scenario, we compute the associated basic reproduction number and we provide conditions for stability. Finally, we formulate an optimal control problem and, by means of simulations, we show that the relative importance of the interventions naturally emerges from the optimal policy.

Keywords: Data-based control, Convex optimization, Output regulation
Abstract: Motivated by perception-based and sensing-based control problems in autonomous systems, this paper addresses the problem of developing feedback controllers to regulate the inputs and the states of a dynamical system to optimal solutions of an optimization problem when one has no access to exact measurements of the system states. In particular, we consider the case where the states need to be estimated from high-dimensional sensory data received only at some time instants. We develop a sampled-data feedback controller that is based on adaptations of a projected gradient descent method and includes neural networks as integral components to estimate the state of the system from perceptual information. We derive sufficient conditions to guarantee (local) input-to-state stability of the control loop. Moreover, we show that the interconnected system tracks the solution trajectory of the underlying optimization problem up to an error that depends on the approximation errors of the neural network and on the time-variability of the optimization problem; the latter originates from time-varying safety and performance objectives, input constraints, and unknown disturbances.

Keywords: Real-time optimal control, Nonlinear predictive control, Stability of nonlinear systems
Abstract: The paper analyzes the closed-loop stability of Dynamically Embedded Model Predictive Control for input-constrained continuous-time nonlinear systems. Given a stabilizing continuous-time optimal control problem, the proposed method performs a discrete approximation to obtain a finite number of optimization variables. The resulting optimization problem is then embedded into a dynamic feedback law that evolves in parallel to the system. Using Input-to-State Stability, it is shown that the dynamic interconnection between the ideal continuous-time model predictive controller and the dynamically embedded solver is asymptotically stable for a sufficiently small discretization step and sufficiently fast solver dynamics. Numerical results, however, highlight a counter-intuitive behavior: as the discretization step decreases, the stability of the closed-loop system tends to deteriorate. This suggests that, although the discretization should be sufficiently accurate to correctly capture the behavior of the system, oversampling the system dynamics may be just as harmful as undersampling.

Keywords: Differential or dynamic games, Networked systems, Sampled-data control
Abstract: We study the steady-state Nash equilibrium-seeking problem for sampled-data games with LTI dynamics and quadratic costs. The key challenge is to guarantee the robust stability and convergence properties of the closed-loop system in the presence of local individual sampling mechanisms assigned to each of the players in the game. This problem is non-trivial due to the unstable behaviors that can arise when sequential control updates (rather than parallel) emerge in the closed-loop system because of the existence of local control triggering mechanisms in each node of the network. To address this issue, we introduce a control framework based on tools from hybrid dynamical systems theory. Our results are illustrated via numerical examples.

Keywords: Differential or dynamic games, Control of interconnected systems, Decentralized control
Abstract: We consider a set of LTI agents subject to constant external disturbances seeking to minimize coupled cost functions in steady-state, modelled as a game-theoretic problem. Using a novel framework, the Nash equilibrium seeking problem has been shown to reduce to the design of a set of decentralized stabilizing controllers. Herein, we consider two such controller designs, based off LMI approaches. The first employs a diagonal stability argument, and the second relies on H-infinity design coupled with the small gain theorem. Applications to sensor networks are provided and the trade-offs between the two methods are discussed.

Keywords: Adaptive control, Differential or dynamic games, Control problems under conflict and/or uncertainties
Abstract: We consider adaptive decision-making problems where an agent optimizes a cumulative performance objective by repeatedly choosing among a finite set of options. Compared to the classical prediction-with-expert-advice set-up, we consider situations where losses are constrained and derive algorithms that exploit the additional structure in optimal and computationally efficient ways. Our algorithm and our analysis is instance dependent, that is, suboptimal choices of the environment are exploited and reflected in our regret bounds. The constraints handle general dependencies between losses (even across time), and are flexible enough to also account for a loss budget, which the environment is not allowed to exceed. The performance of the resulting algorithms is highlighted in two numerical examples, which include a nonlinear and online system identification task.

Keywords: Predictive control, Decentralized control, Networked systems
Abstract: In distributed model predictive control (MPC), the control input at each sampling time is computed by solving a large-scale optimal control problem (OCP) over a finite horizon using distributed algorithms. Typically, such algorithms require several (virtually, infinite) communication rounds between the subsystems to converge, which is a major drawback both computationally and from an energetic perspective (for wireless systems). Motivated by these challenges, we propose a suboptimal distributed MPC scheme in which the total communication burden is distributed also in time, by maintaining a running solution estimate for the large-scale OCP and updating it at each sampling time. We demonstrate that, under some regularity conditions, the resulting time-distributed suboptimal MPC control law recovers the qualitative robust stability properties of optimal MPC, if the communication budget at each sampling time is large enough.

Keywords: Continuous time system estimation, Nonlinear system identification
Abstract: This paper studies the theoretical challenge of linearly representing a class of nonlinearly parameterized estimation problems. Conventional linear regression estimation methods can then be used. For illustration purposes, in our work, the DREM method is chosen to estimate the frequencies for a multi-sinusoidal uncertain signal in a fixed time. Identifying annihilators for signals generated by exponential polynomials with unknown parameters is the first step to accomplishing the linear representation. The annihilators are obtained as time-delay operators constructed with an Ore extension. The fundamental idea is to use the action of these operators on exponential polynomials to obtain the linear regression in nonlinear unknown quantities. We precisely characterize the annihilator as a principal ideal using appropriate Ore extensions, substantially simplifying further calculations. Some examples illustrate our annihilator’s design.

Keywords: Nonlinear system identification, Nonparametric methods, Bayesian methods
Abstract: System identification methods based on Reproducing Kernel Hilbert Spaces (RKHS) have proven to be a valuable tool for the identification of linear time-invariant systems in both discrete- and continuous-time. In particular, unlike most other methods, they enable to systematically confer a priori desirable properties, such as stability, on the estimated models. However, existing RKHS methods mainly target impulse responses and, hence, do not extend well to the nonlinear systems' context. In this work, we propose a novel RKHS-based methodology for the identification of discrete-time nonlinear systems guaranteeing that the identified system is incrementally input-to-state stable (δISS). We model the identified system using a predictor function that, given past input and output samples, yields the output prediction at the next time instant. The predictor is selected from an RKHS using a constrained optimization problem that guarantees its δISS properties. The proposed approach is validated via numerical simulations.

Keywords: Nonlinear system identification, Data-driven control
Abstract: Recently, extended dynamic mode decomposition (EDMD) has become a popular tool for system identification. The EDMD is a data-driven algorithm to identify a linear or bilinear predictor of the nonlinear system based on the Koopman operator. Although linear and bilinear models are suitable for real-time control, these models' prediction accuracy may not be sufficient for problems such as trajectory planning that require long-term prediction. To address this issue, we introduced the nonlinear predictor obtained from a block matrix representation of the bilinear predictor and compared it with the bilinear model. Furthermore, identifying a physically reasonable predictor from the limited number of noisy experimental data is also a challenging problem. In this study, we utilized several simple prior knowledges to achieve stability and sparsity and to identify the model that has correct dependencies between state and control input. We identified the forklift dynamics from experimental data and showed that the obtained model has high prediction accuracy and is suitable for optimal trajectory planning.

Keywords: Nonlinear system identification, LPV system identification, Frequency domain identification
Abstract: Nonlinear (NL) systems are appearing in all engineering applications. Deriving models for these systems is important for instance for prediction and control. The goal of this paper is to estimate models of a class of NL systems via linearization around a time-varying setpoint. By considering one stable trajectory of the NL system, the system can be approximated by a linear parameter-varying (LPV) model around this trajectory. Indeed, this LPV model is the NL system linearized around the trajectory, which we will identify by perturbing the trajectory. After the identification of the LPV model, we reconstruct the NL system by symbolic integration of the estimated LPV coefficients. In this paper, we show that to guarantee the integrability of the LPV coefficients, the parametrization of this LPV model must have a specific structure which we will exploit for its identification. This structure shows that the parameter-varying (PV) coefficients of this LPV are related to each other. Finally, simulation results demonstrate the effectiveness of the proposed identification approach.

Keywords: Nonlinear system identification, Machine learning
Abstract: The SUBNET neural network architecture has been developed to identify nonlinear state-space models from input-output data. To achieve this, it combines the rolled-out nonlinear state-space equations and a state encoder function, both parameterised as neural networks The encoder function is introduced to reconstruct the current state from past input-output data. Hence, it enables the forward simulation of the rolled-out state-space model. While this approach has shown to provide high-accuracy and consistent model estimation, its convergence can be significantly improved by efficient initialization of the training process. This paper focuses on such an initialisation of the subspace encoder approach using the Best Linear Approximation (BLA). Using the BLA provided state-space matrices and its associated reconstructability map, both the state-transition part of the network and the encoder are initialized. The performance of the improved initialisation scheme is evaluated on a Wiener-Hammerstein simulation example and a benchmark dataset. The results show that for a weakly nonlinear system, the proposed initialisation based on the linear reconstructability map results in a faster convergence and a better model quality.

Keywords: Nonlinear system identification, Machine learning
Abstract: In order to make data-driven models of physical systems interpretable and reliable, it is essential to include prior physical knowledge in the modeling framework. Hamiltonian Neural Networks (HNNs) implement Hamiltonian theory in deep learning and form a comprehensive framework for modeling autonomous energy-conservative systems. Despite being suitable to estimate a wide range of physical system behavior from data, classical HNNs are restricted to systems without inputs and require noiseless state measurements and information on the derivative of the state to be available. To address these challenges, this paper introduces an Output Error Hamiltonian Neural Network (OE-HNN) modeling approach to address the modeling of physical systems with inputs and noisy state measurements. Furthermore, it does not require the state derivatives to be known. Instead, the OE-HNN utilizes an ODE-solver embedded in the training process, which enables the OE-HNN to learn the dynamics from noisy state measurements. In addition, extending HNNs based on the generalized Hamiltonian theory enables to include external inputs into the framework which are important for engineering applications. We demonstrate via simulation examples that the proposed OE-HNNs results in superior modeling performance compared to classical HNNs.

Keywords: Stochastic control, Synthesis of stochastic systems, Control under computation constraints
Abstract: This paper is concerned with the coherent quantum linear-quadratic-Gaussian control problem of minimising an infinite-horizon mean square cost for a measurement-free field-mediated interconnection of a quantum plant with a stabilising quantum controller. The plant and the controller are multimode open quantum harmonic oscillators, governed by linear quantum stochastic differential equations and coupled to each other and the external multichannel bosonic fields in the vacuum state. We discuss an interplay between the quantum physical realizability conditions and the Luenberger structure associated with the classical separation principle. This leads to a quadratic constraint on the controller gain matrices, which is formulated in the framework of a swapping transformation for the conjugate positions and momenta in the canonical representation of the controller variables. For the class of coherent quantum controllers with the Luenberger dynamics, we obtain first-order necessary conditions of optimality in the form of algebraic equations, involving a matrix-valued Lagrange multiplier.

Keywords: Estimation and filtering
Abstract: We propose a self-contained and accessible derivation of an exact formula for the n-point correlation functions of the signal measured when continuously observing a quantum system. The expression depends on the initial quantum state and on the Stochastic Master Equation (SME) governing the dynamics. This derivation applies to both jump and diffusive evolutions and takes into account common imperfections of realistic measurement devices. We show how these correlations can be efficiently computed numerically for commonly filtered and integrated signals available in practice.

Keywords: Stochastic control, Synthesis of stochastic systems, Identification for control
Abstract: In variational quantum algorithms (VQAs), the most common objective is to find the minimum energy eigenstate of a given energy Hamiltonian. In this paper, we consider the general problem of finding a sufficient control Hamiltonian structure that, under a given feedback control law, ensures convergence to the minimum energy eigenstate of a given energy function. By including quantum non-demolition (QND) measurements in the loop, convergence to a pure state can be ensured from an arbitrary mixed initial state. Based on existing results on strict control Lyapunov functions, we formulate a semidefinite optimization problem, whose solution defines a non-unique control Hamiltonian, which is sufficient to ensure almost sure convergence to the minimum energy eigenstate under the given feedback law and the action of QND measurements. A numerical example is provided to showcase the proposed methodology.

Keywords: Stochastic control, Realization theory, Complex system management
Abstract: In this paper, we analyze the response of passive quantum linear systems in a non-Markovian environment to single photon input fields. Based on an augmented modelling method for non-Markovian quantum systems, analytic forms of output intensity and output covariance function in steady-state are derived. These results allow us to quantitatively analyze the influence of non-Markovian environment on system response under a framework of linear system theory.

Keywords: Stochastic control, Synthesis of stochastic systems
Abstract: This paper is concerned with quadratic-exponential moments (QEMs) for dynamic variables of quantum stochastic systems with position-momentum type canonical commutation relations. The QEMs play an important role for statistical ``localisation'' of the quantum dynamics in the form of upper bounds on the tail probability distribution for a positive definite quadratic function of the system variables. We employ a randomised representation of the QEMs in terms of the moment-generating function (MGF) of the system variables, which is averaged over its parameters using an auxiliary classical Gaussian random vector. This representation is combined with a family of weighted L^2-norms of the MGF, leading to upper bounds for the QEMs of the system variables. These bounds are demonstrated for open quantum harmonic oscillators with vacuum input fields and non-Gaussian initial states.

Keywords: Distributed optimization for large-scale systems, Multi-agent systems, Graph-based methods for networked systems
Abstract: In this paper, we investigate the problem of distributed load balancing under network capacity constraints, where the participating agents cooperate with the aim of jointly minimizing both the workload disparity among them as well as the overall workload transfer. Classical approaches for asymptotic convergence to the global optimum in a distributed fashion typically assume timely exchange of information between neighboring agents of a given multi-agent system. This assumption is not necessarily valid in practical settings due to non-commensurate (heterogeneous) communication and processing delays that might affect transmissions at different times. More specifically, we consider what effect multiple heterogeneous time-varying delays, among the agents can have on the distributed load balancing problem. We show that the distributed load balancing problem under bounded heterogeneous time-varying delays is globally asymptotically stable, but the rate of convergence is affected. Bounds on the convergence rate are provided with respect to an upper bound on the delays. Simulation examples are provided to show the validity and performance of our theoretical results.

Keywords: Distributed optimization for large-scale systems, Multi-agent systems, Consensus
Abstract: Distributed optimization is an important and practical problem that arose from machine learning, smart grid, and multi-robot systems. In this paper, we propose a zeroth-order gradient tracking method to solve a class of constrained distributed optimization problems with nonidentical feasible sets. We design a more general pseudo-gradient estimation scheme, which includes the existing coordinate descent, discretized gradient descent, and spherical smoothing methods as its special cases. Moreover, we propose pseudo-gradient tracking with projection dynamics to deal with nonidentical feasible set constraints and achieve the optimal solution. We show the proposed algorithm achieves the optimal solution with an O(ln T/√T) convergence rate. Finally, we present an example to demonstrate the effectiveness of the proposed algorithm

Keywords: Distributed optimization for large-scale systems, Convex optimization, Multi-agent systems
Abstract: This paper presents a decentralized algorithm for a team of agents to track time-varying fixed points that are the solutions to time-varying convex optimization problems. The algorithm is first-order, and it allows for total asynchrony, i.e., all communications and computations can occur with arbitrary timing and arbitrary (finite) delays. Convergence rates are computed in terms of the communications and computations that agents execute, without specifying when they must occur. These rates apply to convergence to the minimum of each individual objective function, as well as agents’ long-run behavior as their objective functions change. Then, to improve the usage of limited communication and computation resources, we optimize the timing of agents’ operations relative to changes in their objective functions to minimize fixed point tracking error over time. Simulation results illustrate these developments.

Keywords: Distributed optimization for large-scale systems, Multi-agent systems, Consensus
Abstract: This work studies nonconvex distributed constrained optimization over stochastic communication networks. We revisit the distributed dual averaging algorithm, which is known to converge for convex problems. We start from the centralized case, for which the change of two consecutive updates is taken as the suboptimality measure. We validate the use of such a measure by showing that it is closely related to stationarity. This equips us with a handle to study the convergence of dual averaging in nonconvex optimization. We prove that the squared norm of this suboptimality measure converges at rate mathcal{O}(1/t). Then, for the distributed setup we show convergence to the stationary point at rate mathcal{O}(1/t). Finally, a numerical example is given to illustrate our theoretical results.

Keywords: Distributed optimization for large-scale systems, Event-triggered and self-triggered control, Multi-agent systems
Abstract: This paper addresses a Nash equilibrium (NE) seeking problem for noncooperative games with double-integrator agents who communicate with each other on an unbalanced directed graph. To deal with the unbalance in the consensus term, an auxiliary variable is introduced in the designed distributed NE seeking algorithm and is the estimation of the left eigenvector of the Laplacian matrix associated with the zero eigenvalue. Moreover, an event-triggered broadcasting scheme is proposed to reduce communication loads in the network. It is shown that the proposed communication scheme is free of the Zeno behavior and the asymptotic convergence of the designed algorithm is obtained. Simulation results are demonstrated to validate the proposed methods.

Keywords: Distributed optimization for large-scale systems, Convex optimization
Abstract: The curvature-aided IAG (CIAG) algorithm is an efficient asynchronous optimization method that accelerates IAG using a delay compensation technique. However, existing step-size rules of CIAG are conservative and hard to implement, and the Hessian computation in CIAG is often computationally expensive. To alleviate these issues, we first provide an easy-to-implement and less conservative step-size rule for CIAG. Next, we propose a modified CIAG algorithm that reduces the computational complexity by approximating the Hessian with constant matrices. Convergence results are derived for each algorithm on both convex and strongly convex problems. Numerical experiments on logistic regression demonstrate their effectiveness.

Keywords: Supply chains and networks, Logistics in manufacturing, Complex logistic systems
Abstract: The sharing economy is a new innovative business model for more flexible and sustainable logistics. This paper proposes a stochastic programming model for supply chain network design with truck sharing. The model considers demand uncertainty and optimizes allocation and operation of trucks. A case study with realistic data is presented by applying the model to a supply chain with different collaboration level. A sensitivity analysis of cost parameter for sharing service is also conducted to evaluate the impact on business sustainability. The results show that higher collaboration level successfully copes with demand uncertainty, and improves revenues in the entire supply chain.

Keywords: Logistics in manufacturing, Operations research, Supply chains and networks
Abstract: Resilience refers to the ability to cope with various external risks and shocks; the concept of disaster resilience is an important issue in Japan, where there are many natural disasters. We consider a situation where a large-scale power outage occurred due to a disaster; electrical vehicles (EVs) are used to deliver both of electricity and commodities. In our previous study, we formulated the Electric Vehicle Pickup and Delivery Problem with Time Window (E-PDPTW) for delivery planning of them as a mixed integer programming problem, and we proposed a heuristic method to find feasible solutions in realistic computation time. In this paper, we conduct a case study of the proposed method using realistic data.

Keywords: Smart manufacturing, Cyber-physical production systems
Abstract: Adaptive robotics plays an essential role in achieving truly co-creative cyber physical systems. In robotic manipulation tasks, one of the biggest challenges is to estimate the pose of given workpieces. Even though the recent deep-learning-based models show promising results, they require an immense dataset for training. In this paper, two vision-based, multi-object grasp pose estimation models (MOGPE), the MOGPE Real-Time and the MOGPE High-Precision are proposed. Furthermore, a sim2real method based on domain randomization to diminish the reality gap and overcome the data shortage. Our methods yielded an 80% and a 96.67% success rate in a real-world robotic pick-and-place experiment, with the MOGPE Real-Time and the MOGPE High-Precision model respectively. Our framework provides an industrial tool for fast data generation and model training and requires minimal domain-specific data.

Keywords: Smart assembly, Assembly and disassembly, Smart manufacturing
Abstract: This paper proposes a multi-arm-motion-generation method. Arm robots are expected to substitute workers, especially since multi-arm flexibly is adaptable to diverse product models. However, due to the complexity of the multi-arm motion, many hours of manual motion teaching are required.Therefore, we developed a multi-arm motion-generating method for automatically generating collision-free multi-arm motions with an equal or faster time than manually teaching motion. Our proposed method is a collision-circumvent and stop-avoidance method that takes into account each robot movement vector in time slices. Experiments were conducted to evaluate the proposed method, and the results indicate that the method can generate a short circumvent avoidance trajectory compared with a conventional collision-avoidance method for avoiding all passing areas of robots.

Keywords: Flexible and reconfigurable manufacturing systems, Quality assurance and maintenance
Abstract: Printed circuit board (PCB) measurement and repair is a challenging task that requires experience and expertise to perform. PCB diagnosis and repair shops employ skilled operators to carry out the corresponding measurement tasks using measuring instruments (e.g., oscilloscopes, multimeters) in order to uncover the condition of a particular product. However, these tasks are often repetitive and meticulous, and additionally, the results need to be collected and carefully documented so that the gathered experience regarding the product can be re-used when the next product of the same type arrives into the shop. Nevertheless, the diagnosis of used PCBs is less researched and current flexible automation possibilities are limited. In this paper, a novel visual servoing probe test method and measurement tool are proposed to provide a flexible solution for PCB diagnosis with a higher level of automation. The aim of the approach is to reduce the burden on the operators by carrying out the repetitive measurement tasks and automatically storing the results while leaving the responsibility of measurement profile setup to the human expert. The proposed visual servo system uses manually teached-in measurement points, where template patterns are recorded using cameras, and it is capable of compensating positioning errors in the range of a couple of millimeters. The proof of concept of the proposed method is presented through motherboard measuring experiments, with a 99.7% success rate.

Keywords: Supply chain management , Production planning and control, Operations research
Abstract: Mass customization is the method for effectively postponing the task of differentiating a product for a specific customer until the latest possible point in the supply chain network. We consider the problem of a production planning with customer preferences. In the module production, products are manufactured by combining modules according to customer preferences. We develop a robust optimization model that considers uncertainties in manufacturer and customer decisions, and cost parameters in the supply chain planning. The proposed model is treated as a Stackelberg model in which the leader first optimizes its own objective function and then followers individually find their optimal solutions under the conditions. The problem is formulated as a bilevel production planning problem in which the manufacturer is a leader and multiple customers are followers. The manufacturer's problem is to maximize profit, and customers' problem is to maximize satisfaction with products which are produced by the manufacturer. The problem is formulated as a single-level mixed integer programming problem by adding constraint equations that guarantees the optimality of the lower-level customer satisfaction maximization problem through strong duality to the upper-level constraint equations. Computational results are provided to show the robustness of the proposed model and its ability to achieve efficient production planning.

Keywords: Control of networks, Control under communication constraints, Event-triggered and self-triggered control
Abstract: Most approaches for self-triggered control (STC) of nonlinear networked control systems (NCS) require measurements of the full system state to determine transmission times. However, for most control systems only a lower dimensional output is available. To bridge this gap, we present in this paper an output-feedback STC approach for nonlinear NCS. An asymptotically stable observer is used to reconstruct the plant state and transmission times are determined based on the observer state. The approach employs hybrid Lyapunov functions and a dynamic variable to encode past state information and to maximize the time between transmissions. It is non-conservative in the sense that the assumptions on plant and controller are the same as for dynamic STC based on hybrid Lyapunov functions with full state measurements and any asymptotically stabilizing observer can be used. We conclude that the proposed STC approach guarantees asymptotic stability of the origin for the closed-loop system.

Keywords: Event-triggered and self-triggered control, Observers for linear systems, Learning for control
Abstract: The problem of learning a communication policy is investigated in this paper for the design of an event-triggered observer for discrete-time LTI systems. Firstly, the event-triggered observer problem is formulated as an optimisation problem. The existence of a solution to this problem (communication policy) is investigated and it is verified if this solution still preserves the stability of the estimation error dynamics. Secondly, an algorithm is provided to approximate this optimal solution using neural networks and deep learning. Simulation examples are provided to illustrate the effectiveness of the learned communication policies.

Keywords: Event-based control, Distributed control and estimation, Consensus
Abstract: A drawback of the existing event-triggered distributed observer for a rigid body leader system over jointly connected switching networks is that the upper bounds of two key design parameters were only shown to exist without giving an explicit estimate of the upper bounds. In this paper, by assuming that the communication network is acyclic, we further show that these two design parameters can take any positive value by choosing other parameters appropriately. We will also apply our event-triggered distributed observer to the leader-following consensus problem of multiple rigid body systems and illustrate our design by a numerical example.

Keywords: Event-triggered and self-triggered control, Wireless sensing and control systems, Multi-agent systems
Abstract: Event-based methods carefully select when to transmit information to enable high-performance control and estimation over resource-constrained communication networks. However, they come at a cost. For instance, event-based communication induces a higher computational load and increases the complexity of the scheduling problem. Thus, in some cases, allocating available slots to agents periodically in circular order may be superior. In this article, we discuss, for a specific example, when the additional complexity of event-based methods is beneficial. We evaluate our analysis in a synthetical example and on 20 simulated cart-pole systems.

Keywords: Event-triggered and self-triggered control, Learning for control
Abstract: In this paper, an event-triggering adaptive dynamic programming (ADP) algorithm is developed for the temperature control of large-scale HVAC systems. First, the dynamic model of the system is established by using the conservation of mass and energy, which involves the dynamics of the fan and the cooling coil. Second, the multi-zone temperature control problem is treated as a non-zero-sum game, which requires solving the coupled Hamilton-Jacobian (HJ) equations. Then, the ADP algorithm is employed to solve the HJ equation. Importantly, in order to reduce the network transmission burden of large-scale HVAC systems, the ADP algorithm developed in this paper is based on event-triggering mechanism. Theoretical analysis proves that the tracking error and the neural network weight estimation error are uniformly ultimately bounded. Simulation results verify the effectiveness of the algorithm.

Keywords: Machine learning, Modeling and control of agriculture, Precision agriculture
Abstract: Precision deficit irrigation offers a solution to the increasing global pressure on freshwater resources occasioned by a rising demand for agricultural outputs to support a growing human population. Plant physiological responses to water deficit are describe in terms defining severity of water stress. Implementation of deficit irrigation control strategies capable of achieving the twin goals of maximizing potential yield and minimizing cumulative water consumption requires the identification of water deficit levels corresponding to significant stress thresholds to ensure memory initiation and prevent permanent damage to the crop. In this contribution machine learning approaches are implemented for dynamic identification of water stress thresholds during deficit irrigation of potted maize plants. K-means clustering is initally applied to delineate three zones of water stress described as "no stress", "mild stress" and "high stress" for chronologically segmented data points. Least squares-based polynomial curve fitting is employed to mathematically represent the dynamic progression of stress cluster centroids, with accuracy values ranging between 90 % and 98 %.

Keywords: Design methodologies, Information and sensor fusion, Robots manipulators
Abstract: In recent years, the demand for accurate and delay-free kinematic state estimates, especially regarding acceleration, led to the adoption of inertial sensors placed along the kinematic chain of robotic manipulators. With state-of-the-art signal processing algorithms, arbitrary setups of accelerometers and gyroscopes can be deployed, raising the question of where and how many sensors should be placed for an optimal estimation quality. This paper presents a novel observability-based approach to answer this question independent of a specific estimator implementation. For this purpose, methods for calculating the required sensor measurement ranges and predicting the estimation quality regarding velocity and acceleration are introduced and discussed. The resulting procedure is validated successfully by comparing predicted and actual estimation quality for two example manipulators, indicating that it can provide meaningful aid to a design engineer.

Keywords: Modeling, Mobile robots, Robots manipulators
Abstract: Redundant robots manipulator interested the robotics community in the fact that they can perform additional tasks other than the main one. Assembling them on a mobile platform increases the number of degrees of freedom (DOF) and thus allows them to perform more complex tasks in a larger workspace. However, the increase in the number of degrees of freedom further complicates the establishment of the inverse kinematic model. Indeed, for a given End Effector (EE) pose, several configuration joints can be associated. In this paper, the joint parameterisation of the mobile manipulator (MM) is obtained by clustering the workspace and the configuration space. Thus, the robotic system can be reduced to an equivalent non-redundant system, so that existing conventional methods for solving the inverse kinematics (IK) of a non-redundant manipulator can be applied to derive the IK of the system. The proposed approach is validated by conducting a series of simulations

Keywords: Intelligent robotics, Observer design, Control of switched systems
Abstract: This paper deals with the 2-dimensional paddle juggling where a racket hit a ball vertically and iteratively. The racket rebound model is considered where both the translational and rotational velocities of the ball change at the rebound due to the effect of the racket rubber. The LQ servo controller is derived based on the complete ball transition with the racket rebound model. In addition, a whole observer system to estimate all the ball states is proposed which consists of a continuous one for the free flying ball and a discrete one for the ball rebound. Note that it is essential that the rotational velocity can be only estimated by the discrete observer at the rebound. A numerical simulation is performed in order to verify the effectiveness of the total discrete output feedback controller.

Keywords: Modeling, Guidance navigation and control, Robots manipulators
Abstract: Over recent years, there has been a significant focus on Human Robot Interaction (HRI). The aim of this interaction is to create a mutually beneficial partnership between humans and robots, with robots contributing precision, speed, and force, while humans provide experience, intuition, and high-level management and control strategy understanding. One of the most crucial factors in achieving a successful HRI is the guidance of these robots. This paper proposes the use of reliable Hand Detection based on Long Short-Term Memory (LSTM) and intelligent robot motion planning based on Model Predictive Control Methods (MPC) to detect hand signals and communicate with the robot. By utilizing vision systems, a micro controller-based Edge AI approach, and a wired connection, the reaction speed of the robot can be optimized to maintain a safe separation distance between the human operator and the robot, thus enabling a collaborative and safe environment. The effectiveness and capabilities of this automation framework are validated through a case study.

Keywords: Robots manipulators, Autonomous robotic systems, Mechatronic systems
Abstract: Studies on intelligent robotic manipulation systems have typically focused on the programming efficiency, adaptive control of robotic arms, motion planning of robotic arms, and action diversity of grippers. In this study, a decision tree and visual recognition incorporate into a robotic arm to help it learn complex tasks. This study employed a task tree for the automatic planning of complex tasks, in which a decision-making model was used to generate complex task sets from a pre-built motion dataset in real-time performance. Moreover, the model can analyze the rationality of model steps, introduce new tasks, and perform object analysis. This work applies a support vector machine to identify the state of an object. The model selects a suitable gripper with a rapid gripper switch process by considering the characteristics of the targeting object. This study demonstrated the effectiveness of the proposed approach with suitable intelligence through the assembly task.

Keywords: Robots manipulators
Abstract: The short paper discusses a comparative study of impedance control for experimental hydraulic arms interacting with humans. There are a lot of impedance controls based on different approaches. However, the existing impedance controls have not been compared. First, we review a nominal model of hydraulic cylinder dynamics and the two existing impedance controls for hydraulic arms. Second, via linearization for the short paper, we propose a comparison method of each impedance control. Third, we numerically compare each impedance control. The effectiveness of the comparison method is confirmed.

Keywords: Application of mechatronic principles, Smart structures, Parameter and state estimation
Abstract: Adaptive structures are equipped with sensors and actuators to counteract deformations and vibrations caused by external loads. For railway bridges, active control can be used to reduce vibration amplitudes to extend the service life of the structure, thereby increasing resource and emissions efficiency. Model-based control of bridge structures requires knowledge of the structural state and the external disturbance. This paper compares bridge state estimators (Kalman filters), that get no/ full information about the disturbance or include disturbance models of different complexity in the estimation model with respect to the achievable estimation performance. It is shown that the disturbance cannot be neglected, however it is sufficient to take the average train axle weights as average mass (AM)-moving point load (MPL) model into account, while a more complex disturbance model does not improve the estimation performance. Hence, a state and disturbance estimator based on the AM-MPL model is proposed using an Augmented Kalman filter.

Keywords: Motion control systems, Adaptive control, Disturbance rejection (linear case)
Abstract: In this paper we address the modeling and tracking control problem of an overactuated lift and tilt vertical nanopositioning system. We derive a model based on the rigid body dynamics which adequately describes the first resonance mode of each motion axis. This model is validated with measured data in the frequency domain to illustrate that it adequately reflects the real behavior. We further device an abstract model for the derivation of advanced control strategies. By virtue of the single axis model, three SISO controllers are implemented. The control strategy is accomplished comprising a nominal feedforward and LQG-type controller plus an L1 adaptive augmentation with output feedback. The baseline (or nominal) controller features sufficiently high bandwidth for the mere stabilization, decoupling, and disturbance rejection, while the L1 adaptive component plays a central role for recovering the nominal closed-loop dynamics in the presence of parametric uncertainties w.r.t. the input gain which are quite difficult to handle in the nominal design. The effectiveness and robustness of the proposed control strategy is verified via real-time experiments featuring subnanometer and nanoradian tracking errors which seem to be fully-dominated by the measurement noise.

Keywords: Mechatronic systems, Motion control systems
Abstract: This paper proposes modeling-free inversion-based iterative control (MF-IIC) for a fast steering mirror, which is a dual-input dual-output system to generate 2-D optical scanning trajectories. To explicitly compensate for the cross-coupling motions, the MF-IIC iteratively updates the control inputs by learning from the previous trials based on a 2x2 Jacobian matrix. To operate the MF-IIC without system identification in advance, it estimates the Jacobian matrix including the non-diagonal elements for the cross-coupling dynamics simultaneously during learning. For experiments, 19 Hz and 20 Hz sine waves with an amplitude of 0.65 deg. are used for a Lissajous pattern as a reference trajectory. Experimental results show that a tracking error of 320x10^{-3} degrees when feedback control is used for stability. By additionally combing the proposed MF-IIC, the error is decreased by a factor of 68 to 4.64x10^{-3} degrees. It is smaller than MF-IIC without considering the cross coupling explicitly, which achieves a tracking error of 12.1x10^{-3} deg. These two MF-IIC algorithms are also compared by the resulting sampled trajectories in the spatial domain. For that purpose, maximum dead zone diameter is defined as a diameter of the largest circle without sampled points in a scanning area in this paper. While the MF-IIC without considering the cross coupling explicitly for comparison creates a Lissajous pattern with a maximum dead zone diameter of 74.8x10^{-3} degrees, the proposed MF-IIC achieves a Lissajous pattern with a smaller maximum dead zone diameter of 72.9x10^{-3} degrees, demonstrating its effectiveness for high-quality scanning trajectories.

Keywords: Mechatronics, Mechatronic systems
Abstract: The model uncertaint of hydraulic systems is a key issue in observer-based fault diagnosis. In this research, for diagnosing pressure sensor faults in hydraulic systems, we present the application of adaptive robust observers design. For reducing model uncertainty and increasing the feasibility of the fault diagnosis scheme, this paper uses a robust filter and controlled parameter adaption to deal with model uncertainty. Then, according to the fault system model, the fault magnitude of the pressure sensor fault is estimated and the residual variance can be generated for diagnosing the sensor fault. By experiments, the fault diagnosis scheme is proven efficient, which is based on adaptive robust observers. Then, the fault magnitude can be estimated.

Keywords: Mechatronics, Identification and control methods
Abstract: This paper presents a model-based flux control for a variable reluctance actuator with an electropermanent magnet, used for adaptive zero power gravity compensation in magnetic levitation systems. With the hysteresis of the electropermanent magnet being identified and approximated, tailored current pulses are applied by the model-based flux control to tune the magnetization of the electropermanent magnet. In this way, the resulting stationary reluctance force compensates the gravitational force of the levitated mover mass. Based on the identified hysteresis, a non-linear control law is derived, which is extended by an integrator term to compensate modelling uncertainties. In comparison to the state of the art model-free control, the model-based control increases the force tuning rate by a factor of 14 to 19 N/s and improves the robustness of the experimental system in variable mover positions.

Keywords: Motion control systems, Mechatronic systems, Nonlinear adaptive control
Abstract: Feedforward control can greatly improve the response time and control accuracy of any mechatronic system. However, in order to compensate for the effects of modeling errors or disturbances, it is imperative that this type of control works in conjunction with some form of feedback. In this paper, we present a new adaptive feedforward control scheme for electromechanical systems in which real-time measurements or estimates of the position and its derivatives are not technically or economically feasible. This is the case, for example, of commercial electromechanical switching devices such as solenoid actuators. Our proposal consists of two blocks: on the one hand, a feedforward controller based on differential flatness theory; on the other, an iterative adaptation law that exploits the repetitive operation of these devices to modify the controller parameters cycle by cycle. As shown, this law can be fed with any available measurement of the system, with the only requirement that it can be processed and converted into an indicator of the performance of any given operation. Simulated and experimental results show that our proposal is effective in dealing with a long-standing control problem in electromechanics: the soft-landing control of electromechanical switching devices.

Keywords: Production planning and control, Sustainable Manufacturing, Control of renewable energy resources
Abstract: Incorporating renewable energy sources (RESs) into manufacturing systems has been an active research area in order to address many challenges originating from the unpredictable nature of RESs such as photovoltaics. In the energy-aware scheduling for manufacturing systems, the traditional off-line scheduling techniques cannot always work well due to their lack of robustness with respect to uncertainties coming from imprecise models or unexpected situations. On the other hand, on-line scheduling or rescheduling, which can improve the robustness by using the model and the latest measurements simultaneously, suffer from a high computational cost. This work proposes a hybrid scheduling framework, which combines the advantages of both off-line scheduling and on-line scheduling, to provide a balanced solution between robustness and computational cost. A novel concept of partially-dispatchable state is introduced. It can be treated as a constant in scheduling when the model works well. When the model does not work well, it is triggered as the variable to tune to improve the performance. Such an event-triggered structure can reduce the number of rescheduling and computational costs while achieving a reasonable performance and enhancing system robustness. Moreover, the choice of partiallydispatchable state also provides an extra design freedom in achieving green manufacturing. Simulation examples on a manufacturing system, which consists of a 100-kW solar photovoltaic system, a 10-machine flow shop production line, a 50-kWh energy storage system, a 100-kW gas turbine, and the grid for power supply, demonstrate the validity and applicability of this event-triggered hybrid scheduling (ETHS) framework.

Keywords: Job and activity scheduling, Production planning and control, Logistics in manufacturing
Abstract: Steel production is an energy-intense industry. Besides the production process itself, accompanying logistic processes are characterized by a high energy demand due to the weight of the products. We consider a steel coil storage, where a gantry crane stores and retrieves steel coils of up to 35 tons. An energy-oriented approach for the scheduling of crane operations can help save energy and lower operational costs. In industrial settings, crane operators require swift assistance in sequencing crane operations and assigning storage places for the coils. Therefore, we develop different heuristic approaches for the energy-oriented crane scheduling problem and compare them based on a case study derived from a German steel manufacturer.

Keywords: Job and activity scheduling, Production planning and control, Discrete event systems in manufacturing
Abstract: In a steel hot rolling mill, jobs are scheduled in groups for the production at two different production lines. The process of scheduling involves the selection and the sequencing of jobs as well as the assignment of selectable furnaces to some jobs. An optimal schedule minimizes unproductive times during the production while utilizing necessary retoolings of one production line as best as possible for the production at the other line. The resulting optimization problem is similar to a combination of several traveling salesman problems and orienteering problems. A method is presented to accurately model the alternating production process on both lines, including sequence-dependent setup times, production line retoolings, and the use of an induction furnace with three chambers. Based on this model, unproductive times occurring in a given schedule can be calculated. These times are minimized by a tailored optimization algorithm, which consists of a simulated annealing metaheuristic followed by a local search. The effective application and benefit of this algorithm are demonstrated in a case study.

Keywords: Operations research, Production planning and control, Smart factory
Abstract: Modern manufacturing systems strive to optimize many different key performance indicators. Additionally, mathematical programming-based scheduling models enable the explicit considerations of constraints. However, due to the ever-increasing demand for customized products it is not guaranteed that all constraints can be met at the same time. An approach to circumvent this problem of infeasibility is to replace the constraints by soft-constraints, i.e., to penalize their violation in the objective function. The presence of multiple competing objectives and soft-constraints gives rise to multi-objective optimization problems, where a decision maker has to weigh and balance the different objectives and soft-constraints according to his/her preferences and priorities. Ideally the values of all objectives and soft-constraints should lie within the same order of magnitude, making it easy to weigh them against each other. However, for heterogeneous objectives and soft-constraints with different scales and units this might be challenging. This paper presents an evolutionary algorithm for the optimal parametrization of multi-objective mixed-integer linear programming-based scheduling models. The goal of the evolutionary algorithm is to compute weights for the different terms of the objective which lead to a balanced influence on the overall objective at the optimum. Furthermore, the algorithm is extended by introducing an a priori weighing of the objectives in the fitness function of the evolutionary algorithm. The method is demonstrated on a scheduling problem in which a given set of orders has to be allocated to machines within a manufacturing environment.

Keywords: Advanced planning and scheduling, Modeling of manufacturing operations, Production planning and control
Abstract: We tackle the Flexible Job Shop Problem (FJSSP) subject to availability constraints due to maintenance tasks on machines. Previous studies explored meta-heuristic methods to address this problem and solved small-sized instances generated from classical FJSSP benchmark. We first introduce a new benchmark, comprising more than 2000 instances, for the problem based on the well-known FJSSP benchmarks. A mixed integer model is then developed to solve this problem. We carry out a computational study on the proposed benchmark and analyze its performances. Results show that our model can solve small and medium size instances efficiently. We also observe important variation in complexity depending on the class of instance which makes the proposed benchmark interesting to evaluate our approach and further improve the solving process in future works.

Keywords: Production planning and control, Operations research, Inventory control
Abstract: This study aims to develop an optimal aggregate production planning policy for a reverse logistics system with imperfect state information and chance constraints. The system consists of a forward channel for storing and distributing products and a backward channel for collecting, remanufacturing, or discarding products. The paper presents a chance-constrained linear-quadratic Gaussian optimization model that considers imperfect state information and formulates an equivalent deterministic problem. It also introduces a minimum variance problem to address the uncontrolled variance of the serviceable inventory variable, whose result is an optimal gain to balance the conditional variances of inventory and production over time, making the solution more cost-efficient. The open-loop solution with the minimum variance gain shows its efficiency through a simple example, reducing the total cost of production by controlling the growth of inventory and production variances. Furthermore, this approach offers a practical way for managers to create inventory-production scenarios for decision-making in a reverse logistics system with imperfect state information and chance constraints.

Keywords: Control of renewable energy resources, Optimal operation and control of power systems
Abstract: We consider a day-ahead scheduling problem for resources based on photovoltaic (PV) generation and demand profile predictions.Because the predicted profiles contain uncertainty, the set of profiles is represented as a confidence interval.Giving the predicted profile as a confidence interval, we consider the problem of obtaining ranges for the optimal operating profiles of storage batteries and thermal power plants. This corresponds to finding the region of possible solutions for all parameters representing PV/demand predictions. In order to find the exact solution region efficiently, we focus on the monotonicity of the solution. In particular, we aim to clarify what kind of optimization problem possesses monotonicity. As a first step, we have performed monotonicity analysis in various settings so far. In this study, we show that the problem, where the network structure is taken into account, has monotonicity under conditions with a theoretical proof. We also confirm its practical significance with a numerical example.

Keywords: Control of renewable energy resources, Sliding mode control, Fault-tolerant
Abstract: The increased penetration of wind energy conversion systems into the grid creates many challenges for maintaining grid stability, reliability and efficiency. This paper proposes an approach that integrates the robustness and fast convergence capability of non-singular fast terminal sliding mode control (NFTSMC) with the fast current support of static synchronous compensators (STATCOM) to improve the dynamic stability of wind energy systems during network disturbances. The NFTSMC control approach is designed for the inner current control loop of a STATCOM with voltage source inverter topology. Its main objective is to support both the active and reactive currents and thereby effectively regulate the system’s frequency and bus voltage. The proposed approach was implemented to a WECS-based test microgrid subject to sudden load variations and mismatched disturbances. The obtained results confirmed its effectiveness in properly mitigating dynamic instabilities stemming from grid disturbances and maintaining the voltage and frequency within their rated values. Further comparison between the performance of the proposed NFTSMC and that of a standard SMC approach showed better performance and reduced chattering compared to the standard SMC.

Keywords: Control of renewable energy resources, Modeling and simulation of power systems, Control system design
Abstract: The collector fields of large-scale commercial trough solar power stations often face inhomogeneities in environmental parameters and resistance to flow in the pipe network. Aiming at dealing with the problem of uneven distribution of hydraulic parameters and radiation in large collector fields, this study establishes a heat transfer model and a hydraulic model of the collector field and proposes a control strategy for the heat transfer fluid (HTF) flow in the parabolic trough collector (PTC) loops. Based on the proposed hydraulic model, the loop valve opening is directly calculated and fast control of the loop HTF flow is achieved. Simulation results show that the scheme significantly reduces the differences in HTF temperature at the outlet of the PTC loops.

Keywords: Control of renewable energy resources, Kalman Filtering, Output feedback control (linear case)
Abstract: This work aims to explore effective online torque estimation techniques for tidal turbine control under varying flow conditions. A modified Kalman filter with adaptive features is developed to estimate hydrodynamic torque from a tidal turbine’s available measurements, based on which a modified Newton-Rapson method is employed to calculate the effective flow speed. The rotor speed reference signal required for tidal turbine control can be produced using the estimated torque and flow speed. The Kalman filter model is constructed using on a low frequency lumped parameter model of a horizontal-axis, fixed-pitch, two-bladed variable speed tidal turbine, established from a fully characterised model of a real turbine. The adaptive feature of the Kalman filter allows the tracking of the spatial-temporal variations of the effective flow speed caused by turbulence. Simulation studies are implemented to test the developed algorithm over the full envelope of the flow speeds

Keywords: Control of renewable energy resources, Parameter estimation based methods for FDI, Fault accommodation and Reconfiguration strategies
Abstract: The combined wind speed estimator and tip speed ratio (WSE-TSR) tracking control scheme is widely used to regulate power production for large-scale modern wind turbines. Although very effective, such an advanced control scheme, based on the prior model information, is highly dependent on external measurements. For partial-load region control, the only external information involved is commonly the measured rotor or generator speed. Inaccuracy in such sole measurement results in an unintended turbine operation and might lead to sub-optimal power production and instability. This paper presents a fault-tolerant control (FTC) method, which aims to eliminate the sensor fault effects for modern wind turbine systems. To fulfil this goal, an iterative learning scheme is proposed to detect and estimate the multiplicative sensor fault, on which an adaptive FTC law is formulated such that the effects of the sensor fault are eliminated. Case studies show that the proposed iterative learning FTC method performs well in detecting, estimating, and accommodating the sensor fault under realistic turbulent wind conditions. The advanced wind turbine controller can maintain its control performance even under faulty conditions, preventing further damage to other turbine components and allowing for continuous power production.

Keywords: Control of renewable energy resources, Modeling and simulation of power systems, Nonlinear process control
Abstract: This paper focuses on the utilization of dynamic simulation models in the planning of experiments for control development. The simulation system is a set of models based on the first principles for system level simulation of the complete new TCP-100 research facility at Plataforma Solar de Almerıa (CIEMAT). This new research facility replaced the 32-year-old ACUREX facility with which so many advances in Automatic Control were reached by the research community. The presented models of the parabolic trough collector field (PTC) will be validated with experimental data and the presented simulations are based on the parameter selection from providers’ data sheets and the engineering design project. All state variables are temperatures and input variables include solar radiation, ambient temperature and several setpoints. Several simulation experiments are planned for typical operation days in which the system is operated passing through different operating modes. The nonlinear scaling approach keeps the algorithms unchanged by focusing on the meanings of the measured variables. Variable specific scaling functions are developed for the state and input variables of the facility system level model extended with several temperature differences. The functions of the irradiance do not change which means that also the indicator of the cloudiness remains the same. The preliminary simulation experiments in a limited set of subsystems will be extended before going to the test campaigns with the new facility.

Keywords: Modeling and simulation of power systems, Smart grids, Control system design
Abstract: In this paper, we present finite-dimensional port-Hamiltonian system (PHS) models of a gas pipeline and a network comprising several pipelines for the purpose of control design and stability analysis. Starting from the partial differential Euler equations describing the dynamical flow of gas in a pipeline, the method of lines is employed to obtain a lumped-parameter model, which simplifies to a nonlinear third-order PHS. Parallels between gas networks and power systems are drawn by showing that the obtained pipeline PHS model has the same pi-representation as electrical transmission lines. Moreover, to assist future control design, additional passivity properties of the pipeline PHS model are analysed and discussed. By comparing the proposed PHS models against other models in a standard simulation, we show that the simplifying assumptions have no material effect on the model fidelity. The proposed pipeline and network models can serve as a basis for passivity-based control and analysis while the power system parallels facilitate the transfer of existing methods.

Keywords: Modeling and simulation of power systems, Smart grids, System identification and modelling
Abstract: Using hydraulic models in control of district heating networks can increase pumping efficiency and reduce sensitivity to hydraulic bottlenecks. These models are usually white-box, as they are obtained based on full knowledge of the district heating network and its components. This type of model is time-consuming to obtain, and might differ from the actual behavior of the system. In this paper, a method is proposed to obtain a grey-box hydraulic model for tree-shaped district heating systems: hydraulic parameters are estimated based on pressure measurements in only two locations. While previous works only estimate parameters related to pressure losses in pipes, this work also includes customers valves in the grey-box model structure, an important inclusion for control-oriented applications. Finally, a numerical example illustrates the proposed method on a small district heating network, showing its ability to obtain an accurate model on the basis of noisy measurements.

Keywords: Optimal operation and control of power systems
Abstract: We propose an optimal operation controller for an electro-thermal microgrid. Compared to existing work, our approach increases flexibility by operating the thermal network with variable flow temperatures and in that way explicitly exploits its inherent storage capacities. To this end, the microgrid is represented by a multi-layer network composed of an electrical and a thermal layer. We show that the system behavior can be represented by a discrete-time state model derived from DC power flow approximations and 1d Euler equations. Both layers are interconnected via heat pumps. By combining this model with desired operating objectives and constraints, we obtain a constrained convex optimization problem. This is used to derive a model predictive control scheme for the optimal operation of electro-thermal microgrids. The performance of the proposed operation control algorithm is demonstrated in a case study.

Keywords: Optimal operation and control of power systems, Power systems stability
Abstract: We study input and state constrained inverse optimal control problems starting from a stabilizing controller with a control Lyapunov function, where the goal is to make the controller an explicit solution of the resulting constrained optimal control problem. For an appropriate cost design and initial states for which a sublevel set of the Lyapunov function is contained in the state constraint set and the initial input lies on an ellipsoid inside the input constraint set, we show that the stabilizing controller solves the constrained optimal control problem. Compared to the state-of-the-art, we avoid solving nonlinear optimization problems evaluated pointwise, i.e., for every state, or in a repetitive fashion, i.e., at each time step. We apply our theoretical results to study the angular droop control studied in (Jouini et al., 2022) of an inverter-based power network. For this, we accommodate the constraints on the angle and power generation and exemplify our approach through a two-inverter case study.

Keywords: Smart grids, Analysis and control in deregulated power systems, Optimal operation and control of power systems
Abstract: We propose a game-theoretic market mechanism for energy balancing in a real-time market and formulate the competition among energy consumers as a Generalized Nash Game (GNG). In this framework, the supply function based bidding method is adopted to mitigate the market power of active energy consumers. Moreover, the physical constraints are incorporated to guarantee the secure operation of the distribution network. To steer consumers to the Generalized Nash Equilibrium (GNE) of this game, existing studies usually require participants to share full or partial private information which may not be appropriate for the practical implementation. In this regard, we design a preconditioned forward-backward based algorithm with provable convergence, by which a market participant only needs to share limited non-private information with others.

Keywords: Smart grids, Extremum seeking and model free adaptive control, Real time optimization and control
Abstract: This paper proposes a Multi-stage Home Energy Management System (MS-HEMS) for power demand distribution among the Photovoltaic system (PV), the Energy Storage System (ESS), and the Electrical Power Grid (EPG). MS-HEMS consists of two layers: the Anticipative layer (AL) and the reactive layer (RL). The AL employs Particle Swarm Optimization (PSO) for day-ahead energy management based on weather and energy consumption forecasts; the RL includes an Extremum-Seeking Controller (ESC) that determines the ideal power setpoint of each source in real-time, compensating for prediction uncertainties and calculation time horizons. The optimization problem considers the energy bill, Peak to Average Ratio (PAR), and battery degradation cost. The proposed MS-HEMS is highlighted using predicted and actual measurements and increased the energy bill gain by 10.8% while reducing the PAR by 56.1% compared to the offline approach (OF-HEMS).

Keywords: Reinforcement learning and deep learning in control, Machine learning in modelling, prediction, control and automation, Traffic control systems
Abstract: Traffic control is essential to reduce congestion in both urban and freeway traffic networks. These control measures include ramp metering and variable speed limits for freeways, and traffic signal control for urban traffic. However, current traffic control methods are either too simple to respond to complex traffic environment, or too sophisticated for real-life implementation. In this paper, we propose an adaptive parameterized control method for traffic management by using reinforcement learning algorithms. This method takes advantage of the simple structure of parameterized state-feedback controllers for traffic; meanwhile, a reinforcement learning agent is employed to adjust the parameters of the controllers on-line to react to the varying environment. Therefore, the proposed method requires limited real-time computational efforts, and is adaptive to external disturbances. Furthermore, the reinforcement learning agent can coordinate multiple local traffic controllers when adjusting their parameters. The method is validated by a numerical case study on a freeway network. Results show that the proposed method outperforms conventional controllers when the system is exposed to a changing environment.

Keywords: Neurodynamic optimization and adaptive dynamic programming, Machine learning in modelling, prediction, control and automation, Quantized systems
Abstract: Based on our recent research on neural heuristic quantized systems, we propose an emulation problem consistent with our recently introduced {em neuromimetic} paradigm. This optimal quantization problem can be solved with model predictive control (MPC) by deriving the conditions under which the quantized system can simultaneously guarantee (asymptotic) stability and emulation given dynamical system by optimizing a Lyapunov-like objective function. Taking inspiration from neurobiology, the neuromimetic model features large numbers of discrete inputs that collectively produce stable motions that emulate the behavior of a continuous system. The emulation is produced by solving an optimization problem involving integer variables. The approach in the paper begins by performing the optimization using model predictive control (MPC) and then using a neural network to train a model using the data generated in this process. Complexity is reduced by applying Fincke and Pohst's sphere decoding algorithm to narrow down the search for the optimal solution.

Keywords: Multi agent systems, Real-time algorithms, scheduling, and programming
Abstract: The control and path planning of multiple drones in a 3D space with obstacle avoidance is a complex challenge, subject to ongoing research. In this paper we propose an optimal control approach, with a baseline controller and filter running on the drones, and a prediction based optimization algorithm running remotely, as a supervisory controller. The supervisor is responsible for calculating the minimal deviation from the trajectory given by the baseline controllers, such that the obstacle is avoided. The approach is tested in simulations with nonlinear drone models in a realistic setting with noise and transmission delays.

Keywords: Neurodynamic optimization and adaptive dynamic programming, Reinforcement learning and deep learning in control, Multi agent systems
Abstract: This paper presents a distributed adaptive formation control for large-scale multi-agent systems(LS-MAS) that addresses the ``Curse of Dimensionality". A novel hybrid game theoretic algorithm that effectively integrates the mean field, Stackelberg, and cooperative game seamlessly has been developed. In particular, LS-MAS has been separated into a multiple number of groups, each has one group leader and a significant amount of followers. Next, a cooperative game is utilized for the inter-group formation of leaders, the mean-field game is adopted for intra-group followers, and a Stackelberg game is connecting the leader and followers in same group. A hybrid reinforcement learning algorithm learns the solution of the hybrid game optimal distributed formation. It comprises a multi-actor-critic to obtain the optimal formation control among inter-group leaders in a distributed manner, and a Stackelberg game-based actor-critic-mass algorithm to obtain the followers' adaptive formation control. Lastly, to demonstrate the efficacy of the proposed approaches, numerical simulations have been conducted.

Keywords: Reinforcement learning and deep learning in control, Traffic control systems, Multi agent systems
Abstract: Traffic signal control is important in intelligent transportation system, of which the cooperative control is difficult to realize but yet vital. Popular methods for solving this problem are based on multi-agent reinforcement learning (RL), in which function approximator, e.g., different kinds of neural network play a critical role. In this paper, we propose a multi- agent actor-critic RL framework with global value function and local policy function, for which the piecewise linear neural network, named biased ReLU (BReLU) is used as the function approximator. The reason for doing this is two-fold. First, it has been proved in the control literature that minimizing (maximizing) a piecewise linear function over a polyhedron yields piecewise linear solutions. Second, the BReLU neural network can provide a more accurate approximation than the traditional ReLU neural network when they have similar network structures. The proposed method is evaluated on the Simulation of Urban Mobility (SUMO) environment compared with two benchmark traffic signal control methods. The simulated results illustrate the proposed algorithm can coordinate the signal control between different intersections, achieve lower and more sustainable intersection delays on the whole traffic network.

Keywords: Multi agent systems, Data fusion and data mining in control
Abstract: This paper presents an active sensor fusion technique for multiple mobile agents (robots) to detect an unknown number of static targets at unknown positions. To process and fuse sensor measurements from the agents, we use a random finite set formulation with an iterated-corrector probability hypothesis density filter. Our main contribution is to introduce two different multi-agent planners to quickly find the targets. The planners make greedy decisions for the next state of each agent by maximizing an objective function consisting of target refinement and exploration components. We demonstrate the performance of our approach through a series of simulations using homogeneous and heterogeneous agents. The results show that our framework works better than a lawnmower baseline, and that a centralized version of the planner works best.

Keywords: Data-based control, Observer design, Nonlinear observers and filter design
Abstract: Digital twins have emerged in recent years as software-based simulation tools that can mirror the behavior of complex dynamical systems. These tools often contain internal representations of a system state that may not be readily available online. In this paper, we develop a data-driven moving horizon estimator for estimating digital twin states only from online measurements. Our framework combines the high expressiveness of deep autoencoders with a moving horizon state estimator that accurately predicts the internal state of the digital twin without complete knowledge of the true system dynamics. We demonstrate that our approach outperforms extended and Koopman Kalman filter solutions on a benchmark reverse van der Pol oscillator example.

Keywords: Data-based control, Output feedback control (linear case), Learning for control
Abstract: While the optimization landscape of policy gradient methods has been recently investigated for partially observed linear systems in terms of both static output feedback and dynamical controllers, they only provide convergence guarantees to stationary points. In this paper, we propose a new policy parameterization for partially observed linear systems, using a past input-output trajectory of finite length as feedback. We show that the solution set to the parameterized optimization problem is a matrix space, which is invariant to similarity transformation. By proving a gradient dominance property, we show the global convergence of policy gradient methods. Moreover, we observe that the gradient is orthogonal to the solution set, revealing an explicit relation between the resulting solution and the initial policy. Finally, we perform simulations to validate our theoretical results.

Keywords: Data-based control, Predictive control, Linear systems
Abstract: In this paper, we propose a data-driven economic model predictive control (EMPC) scheme with generalized terminal constraint to control an unknown linear time-invariant system. Our scheme is based on the Fundamental Lemma to predict future system trajectories using a persistently exciting input-output trajectory. The control objective is to minimize an economic cost objective. By employing a generalized terminal constraint with artificial equilibrium, the scheme does not require prior knowledge of the optimal equilibrium. We prove that the asymptotic average performance of the closed-loop system can be made arbitrarily close to that of the optimal equilibrium. Moreover, we extend our results to the case of an unknown linear stage cost function, where the Fundamental lemma is used to predict the stage cost directly. The effectiveness of the proposed scheme is shown by a numerical example.

Keywords: Linear systems, Time-invariant systems, Data-based control
Abstract: The paper proposes a system representation formed by a minimal collection of sufficiently long restricted trajectories generated by an observable discrete time LTI system. Conditions are given under which such a collection is a system representation and also an exhaustive parametrization of these representations is provided. These can be also interpreted as a generalized persistency condition which complements the results encountered for the controllable case. In terms of the proposed representation some system properties are investigated and a controllable--autonomous decomposition is given. Finally it is shown how the representation associated to the inverse system, to the parallel and cascade connection, respectively, can be derived.

Keywords: Data-based control, Sampled-data control
Abstract: This paper is concerned with the formal synthesis of safety controllers for partially-observable continuous-time polynomial-type systems with unknown dynamics. Given a continuous-time polynomial-type estimator with a partially-unknown dynamic and a known upper bound on the estimation accuracy, we propose a data-driven approach to compute a polynomial-type controller ensuring safety of the system. The proposed framework is based on a notion of so-called control barrier functions and only requires a single output trajectory collected from the system and a single state trajectory collected from its estimator. We show the application of our technique by synthesizing a safety controller for a partially-observable jet engine with unknown dynamics.

Keywords: Data-based control, Machine learning in modelling, prediction, control and automation, Data-driven optimal control
Abstract: Bayesian optimization (BO) is a widely used data-driven method for the global optimization of black-box objective functions with noise. A key component of BO is surrogate model that is a probabilistic model used for approximating black-box objective functions. However, classical BO methods often utilize Gaussian processes with stationary covariance functions as surrogate models, which is difficult to accurately model complex systems with rapidly oscillating target functions, and therefore deteriorates the performance of BO. To this end, a Bayesian optimization approach based on deep kernel learning (DKL) is investigated for the optimization of complex objective functions with large oscillations, in which a Gaussian process based on deep kernel learning is utilized to capture the characteristics of data with non-stationary and hierarchical covariance functions via deep neural networks. The optimization performance of the DKL-based BO approach with different network structures is quantified and compared with a state-of-the-art BO method based on Gaussian process under the scenarios of different noise levels.

Keywords: Diagnosis, Linear systems, Time-invariant systems
Abstract: This paper proposes a new method for the computation of guaranteed minimal detectable and isolable faults of set-based active fault diagnosis for multiplicative actuator faults. Compared with our previous work, the proposed method in this paper has the important advantage of low computational complexity. In particular, by handling the coupling of faults and inputs and using an objective function describing the minimal detectable and isolable faults, a new two-step optimization method is proposed to analyze and compute the minimal detectable and isolable faults. At the end of this paper, two examples are used to show the effectiveness of the proposed method and the advantage of low computational complexity.

Keywords: Fault-tolerant, Linear systems, Robust control (linear case)
Abstract: We propose dependability management of control systems via controller design under the authorization by international standards. We can obtain a controller achieving a given target value of availability and the optimal performance in the normal operation. Furthermore, we show quantitatively the existence of the trade-off between the availability target and control performance first time ever. This is meaningful for the further developments in fault-tolerant control.

Keywords: Fault-tolerant, Adaptive and robust control of automotive systems, Lyapunov methods
Abstract: A new fault-tolerant longitudinal control allocation is developed in this paper in order to distribute the different torques considering that the vehicle involved is an electrical four wheel driving vehicle. Its purpose is to keep an acceleration set-point given by the driver and to reckon with the wheels slip that can occur in an off-road terrain. The method is based on a constrained dynamic pseudo-inverse allocation.

Keywords: Fault-tolerant, Adaptive control, Constrained control
Abstract: This paper proposes a robust adaptive attitude control law for handling a class of actuator faults in the presence of various challenges, such as time-varying inertia, attitude constraints, and input saturation. The proposed approach uses cone angles to represent orientation errors, which are constrained within a performance function to ensure desired transient and steady-state behaviour during reference tracking. Input saturation is approximated using a smooth hyperbolic tangent function. The Nussbaum gain technique is used to handle the unknown control coefficients, which guarantees uniformly ultimately bounded stability in the presence of uncertainties and disturbances. The paper also proposes a norm-based disturbance approximation to estimate the total uncertainty during the Lyapunov analysis. Numerical simulations demonstrate the effectiveness of the proposed control law.

Keywords: Fault-tolerant, Data-driven robust control, Adaptive control
Abstract: In this paper, we present a heuristic method for adapting a networked linear feedback controller to improve its robustness to timing complications, such as long delays, aborted computations, and dropped packets. The core concept of the approach is to log successful sampling and actuation events and then, at discrete time-points, use non-convex parametric optimization to improve the expected performance of the controller under the assumption that the future timing behavior will be similar to the current one. To reduce the time complexity of the optimization algorithm, automatic differentiation is integrated for efficient gradient descent. The approach is evaluated on a physical ball and beam plant, where both the controller and optimization algorithm can be located in the Cloud.

Keywords: Fault-tolerant, Linear systems, Switching stability and control
Abstract: In this paper, as a support technology for safely performing preventive maintenance of LTI control systems, we propose a safe shutdown process for a maintenance-object subsystem. The proposed process is based on the ``fail-soft'' concept that the subsystem gradually makes its way toward a complete stoppage and the overall control system is lead to a safe situation by using the remaining function in the overall system including the subsystem. The discussion in this paper can be applied to a return process to the normal operation.

Keywords: Cooperative navigation, Multi-vehicle systems, Multi-agent systems
Abstract: This paper studies formation control of multi-lane platoons of vehicles, endowed with local positioning capabilities, under the influence of noises and inconsistencies of inter-vehicle relative position measurements. We propose a distributed two-level control framework that consists of a high-level relative position based distributed formation control scheme and low-level individual dynamic controllers of the platoon vehicles. The proposed formation control scheme utilizes deadzone based switching for robustness against sensor noises and adaptive longitudinal controllers for enabling the platoon of heterogeneous vehicles to track the desired platoon velocity.

Keywords: Cooperative navigation, Multi-vehicle systems, Motion control
Abstract: Cooperative adaptive cruise control (CACC) is a control method that enables close vehicle following with string stable behavior, i.e., disturbances are not amplified through the vehicle string. Consequently, the technique contributes to increased road throughput and safety. General practice in CACC design is to describe the longitudinal vehicle dynamics by a simple first order model. For certain vehicles, however, this model has to be extended by an input delay to better represent the longitudinal behavior. In this paper, we show that the performance of a class of heterogeneous CACC controllers greatly deteriorates when the ego vehicle has a delay in the driveline. The minimum required string stable timegap increases to such a degree, that the controller is no longer suitable for close vehicle following. By implementing a Smith predictor in the control scheme, the delay is compensated, restoring the string stability properties. However, the original vehicle following objective is no longer fulfilled with the inclusion of the Smith predictor. By adopting an alternative timegap, the original control objective is restored for steady state situations, enabling CACC for heterogeneous platoons with actuator delays.

Keywords: Multi-vehicle systems, Cooperative navigation, Autonomous vehicles
Abstract: A distributed coordination method for solving multi-vehicle lane changes for connected autonomous vehicles (CAVs) is presented. Existing approaches to multi-vehicle lane changes are passive and opportunistic as they are implemented only when the environment allows it. The novel approach of this paper relies on the role of a facilitator assigned to a CAV. The facilitator interacts with and modifies the environment to enable lane changes of other CAVs. Distributed MPC path planners and a distributed coordination algorithm are used to control the facilitator and other CAVs in a proactive and cooperative way. We demonstrate the effectiveness of the proposed approach through numerical simulations. In particular, we show enhanced feasibility of a multi-CAV lane change in comparison to the simultaneous multi-CAV lane change approach in various traffic conditions generated by using a data-set from real-traffic scenarios.

Keywords: Teleoperation, Guidance, navigation and control of vehicles
Abstract: In vehicle teleoperation, reducing instability of driving caused by communication delay is one of important technical problems. This study proposes a passivity-based vehicle teleoperation system that compensates for the delay of visual information presented to the driver. The proposed method applies wave variable transformation to steering angle and yaw rate, and presents modified yaw rate to the driver by altering the viewing direction of camera images. This effectively makes the driver feel that the vehicle responds to his/her steering operation without delay, and consequently over-steering is greatly reduced. The results of experiments using a driving simulator showed that the proposed method greatly improves lateral stability of driving.

Keywords: Intelligent Transportation Systems, Modelling and control of road traffic networks
Abstract: The perimeter control is effective in alleviating traffic congestion. However, previous studies about perimeter control based on macroscopic fundamental diagram didn’t consider the influences of the system uncertainty even though this uncertainty may lead to significant congestion. This paper presents a two-region perimeter control method based on a risk-averse model predictive control considering uncertainty. The advantage of this paper is that we design a risk index for uncertainty and design a controller which takes the risk index as an optimization objective. A scenario tree is used to model the uncertainty of the two-region dynamic equations. Average value at risk mapping is employed to calculate the systematic risk due to uncertainty. Then, the optimization objective of a multi-stage risk is used in MPC based on the scenario tree. Finally, the multi-stage risk is formulated as a solvable form. Simulation results show that the proposed perimeter control method reduces the total travel time of the two-region road network with uncertainty and significantly reduces traffic congestion compared to the stochastic model predictive control method.

Keywords: Autonomous vehicles, Autonomous systems, Trajectory tracking and path following
Abstract: We present an analytical method to estimate the collision probability of motion plans for autonomous agents with discrete-time dynamics operating under Gaussian motion and sensing uncertainty. Motion plans generated by planning algorithms cannot be perfectly executed by autonomous agents in reality due to the inherent uncertainties in the real world. Estimating end-to-end collision probability is crucial to characterize the safety of trajectories and plan risk-optimal trajectories. In this work, we derive upper and lower bounds for end-to-end collision probability of motion plans using results from probability theory including the inequalities of Hunter, Kounias, Frechet, and Dawson. Using a ground robot navigation example, we demonstrate that our method is considerably faster than the naive Monte Carlo sampling method and the proposed bounds are significantly less conservative than Boole’s bound commonly used in the literature.

Keywords: Bio-signals analysis and interpretation, Biomedical and medical image processing and systems
Abstract: Electrical impedance tomography (EIT) is a promising imaging technology. It is portable, user-independent, and low cost (compared to other imaging technologies). There are two main approaches for EIT equipments: dynamic and absolute images. Absolute images are technologically more difficult and are limited by hardware. Phase information is required to create permittivity images. This is an ongoing research with the objective of evaluating the phase information. A new architecture with up to 64 electrodes is proposed. Each electrode has its own Howland current source and microcontroller. The proposed hardware consists of three main modules: the electrode hardware (with Howland current source and microcontroller); the demodulator algorithm; and, the electrode synchronizer. Some results are presented with the electrode hardware that demonstrate data acquisition with two electrodes. Tests with the demodulator algorithm were also presented. Simulated tests with the designed electrode synchronizer were performed, and the implementation remains, and it is considered a future work.

Keywords: Medical imaging and processing, Developments in measurement, signal processing, Decision support and control
Abstract: Electrical Impedance Tomography (EIT) is a commonly used imaging technique for monitoring respiration on the bedside and it might have the potential for monitoring lung perfusion. Several signal processing approaches have been developed to separate respiration and perfusion. In this contribution we investigated whether different image reconstruction algorithms influence the separation results provided by the harmonic analysis approach. We compared the algorithms used by Dräger, the Gauss-Newton method with different regularizers as well as the GREIT algorithm. The comparison was carried out using a retrospective EIT dataset from a COVID-19 patient. The results gave insight that the harmonic analysis separation approach is dependent on the reconstruction algorithms. Both, the separation of the perfusion and the separation of the respiration showed differences between the reconstruction algorithms when carried out pixel-wise. On the other hand, the separations carried out on the global impedance only showed marginal differences for the separated perfusion.

Keywords: Medical imaging and processing
Abstract: Mammography is currently considered the gold standard for breast screening, despite painful breast compression, invasive radiation exposure and excessive infrastructure and personnel requirements, contributing to its inequity. Furthermore, sensitivity and specificity of mammography are often overstated by studies utilising flawed methodology. This paper critically reviews the literature reporting on methods used to calculate diagnostic accuracy of mammography. Values for sensitivity, specificity and receiver operator characteristic (ROC) curve area (AUC) are presented by averaging results from studies inclusive of a comparison to another breast screening modality. The result is mammography sensitivity of 60%, specificity of 80%, AUC of 0.73. These values should be used when assessing diagnostic performance of other breast screening technologies.

Keywords: Biomedical and medical image processing and systems, Decision support and control, Medical imaging and processing
Abstract: Analysing laparoscopic videos, particularly for surgical tool classification and localisation, has attained interest in the field of surgical data science since they represent an extensive information source. However, the difficulties for acquiring and labelling these videos have led to paucity of labelled datasets. Consequently, the progress of developing robust and generalised surgical tool detection models was slowed down, and translating these models into the medical field was hindered. In this work, supervised surgical tool classification and weakly-supervised tool localisation in laparoscopic videos were addressed. A base convolutional neural network (CNN) model was adapted to perform both tasks by incorporating multi-map localisation layers. Squeeze-and-excitation modules were added to the CNN to enhance the ability of the model to generate better focused informative features. Additionally, features at multiple stages of the CNN were combined and fused in a batch normalisation layer to enhance model generalisability. The proposed model was evaluated on the popular Cholec80 dataset. Experimental results of 94.1% mean average precision for tool classification and 70.1% F1-score for tool localisation revealed the ability of the model to learn better features for both tasks. The proposed approach showed the advantages of integrating attentions modules and multi-stage features fusion technique for surgical tool classification and localisation.

Keywords: Biomedical and medical image processing and systems, Biomedical system modeling, simulation and visualization, Medical imaging and processing
Abstract: Electrical impedance tomography (EIT) is an imaging technology but suffers greatly from the ill-posed inverse problem when reconstructing an image, which is mainly caused by the high degrees of freedom and the relatively large measurement noise. The use of discrete cosine transform (DCT) to cluster finite elements has been proposed to reduce the degrees of freedom in inverse computations. However, blurred anatomical alignment and artifacts still present challenges to the interpretation of EIT images. Incorporating prior information into the reconstruction process has been reported to enhance the quality of EIT images. In this contribution, we propose the use of a patient-specific structural prior mask for the DCT-based EIT algorithm. We evaluate the influence of this mask on simulation models with varying ventilation statuses. Our results demonstrate that the structural prior mask preserves the morphological structures of the lungs and avoids blurring of the solution, thereby facilitating EIT image interpretation for clinicians.

Keywords: Medical imaging and processing, Biomedical and medical image processing and systems, Decision support and control
Abstract: The segmentation of brain tissues based on MRI data represents a widely investigated problem in medical image analysis. Infant brain tissues in the so-called isointense phase of development, around the age of 6 to 9 months, are difficult to segment due to the inherent myelination and maturation processes, which manifest in grey and white matter pixel intensity distributions severely overlapping in both T1-weighted and T2-weighted MRI data. This paper introduces two different U-net architectures, one with 2D convolution and this other with 3D convolution, and employs them in infant brain tissue segmentation based on multi-spectral MRI data. The proposed methods are trained and tested on the training data set of the iSeg-2017 challenge. All records were initially fed to histogram alignment that was independently performed on each data channel. Nine records were used for training and one record for testing in each evaluation round, but all ten records took their turn as testing data. Statistical accuracy indicators were employed to assess the segmentation accuracy of both network models. The U-net architecture preforming 3D convolution obtained the better segmentation, achieving 91.8% average rate of correct decisions, and 95.0%, 91.5% and 89.8% Dice similarity score for cerebro-spinal fluid, grey matter, and white matter tissues, respectively.

Keywords: Digital twins for manufacturing, Intelligent manufacturing systems, Smart manufacturing
Abstract: Both Digital Twins (DTs) and Artificial Intelligence (AI) are major elements in the Industry 4.0 paradigm. However, while DTs embody the Cyber Physical Systems (CPS) vision of connected and digitally controlled world, AI integration in DT has no clear and agreed upon framework. In this paper we focus on DTs in Industry 4.0 environments and characterize the implementation of AI in these DTs. In the first part of the paper, we review the relevant literature. In the second part of this paper, we discuss the challenges for integrating AI in DT. Finally, we identify both practical and research potential of this integration and discuss their advantages and shortcomings, as well as difficulties that are expected in their deployment.

Keywords: Human-centric manufacturing, Intelligent manufacturing systems, Human-automation integration
Abstract: The shift from Industry 4.0 to Industry 5.0 has led to a greater focus on workers' needs in the workplace. Collaborative robots have been introduced to promote a fair division of tasks and reduce physical and mental strain on workers. However, there is a lack of research on how to implement human-centered task allocation. This study proposes a model for multi-objective task allocation, including minimizing makespan, energy expenditure, and mental workload. The study also suggests a method for evaluating mental workload. Results show that the strictness of task sequence affects makespan and energy expenditure, and a new constraint related to idle times is proposed. The optimal level of worker saturation is one that minimizes makespan while minimizing increases in energy expenditure and mental workload.

Keywords: Digital twins for manufacturing, Discrete event systems in manufacturing
Abstract: Digital Twin (DT) is one of the promising digital technologies being developed at present to support digital transformation. Successful DT execution needs real-time access to collected data. However, real-time connection and synchronization are a significant difficulty, owing to the physical environment's unique characteristics such as variability, uncertainty, the different scales of the physical and virtual spaces, and the data generated continuously by various entities. Prior research has been conducted to address the DT synchronization problem in the literature. However, there is a lack of standard ap-proaches or mechanisms for digital twins at the field level in manufacturing systems. Hence, this paper aims to present a reliable field synchronization mechanism for DT of manufacturing systems for real-time data.

Keywords: Flexible and reconfigurable manufacturing systems, Smart manufacturing, Manufacturing plant control
Abstract: This paper focuses on the parts routing problem in a reconfigurable manufacturing plant, in presence of potential faults and uncertainty on the job scheduling and duration. The plant is modeled as a directed graph, where the nodes represent either transportation modules or machines, and the edges represent the allowed transitions between adjacent nodes. The parts move across the plant along predefined sequences of nodes, therefore the system state tracks the progress of the parts along such sequences and the control inputs are the transitions to be activated to command the parts movement. Provided the sequences, the proposed method automatically generates feedback control rules for deadlock avoidance, which are employed by a path following strategy to compute the suitable control inputs, complying with given temporal-logic constraints and avoiding deadlock states. Additionally, the approach is extended to deal with faults affecting the transportation modules via the selection of new feasible sequences and the online reconfiguration of the system state. Finally, the proposed approach is tested in high-fidelity simulations, showing high computational efficiency and throughput.

Keywords: Digital twins for manufacturing, Advanced planning and scheduling, Manufacturing plant control
Abstract: The interest in enhancing decision-making using Digital Twins in production systems is increasing in the recent years. However, most of the scientific contributions in the field lack of generality as they focus specific problem instances, overburdening the search for a common thread. To help overcoming this issue, the authors propose a conceptual framework in order to analyze the role of Digital Twins in applications in different production environments, with specific concern on production scheduling and control. The analysis is achieved by responding to three main questions, which focus on the application, the integration, and the functionality of the Digital Twin. The framework is applied to describe six relevant use cases of Digital Twins within this research area.

Keywords: Digital twins for manufacturing, Human-centric manufacturing, Cyber-physical production systems
Abstract: The digital twin of human operators is an important aspect of future cyber-physical production systems. Among the various usage of this twin, the possibility to assess the health of the operator, especially fatigue and musculoskeletal disorders, is a major stake for future organizations. If this problem is often dealt with in the case of repetitive operations, mostly through an ergonomic perspective, it is much more difficult to evaluate the impact of non-repetitive operations on the health of the operators. This study aims at providing a first step to this objective, by exhibiting the correlation between ergonomic evaluations of operations difficulty with the MUSKA MSD methodology and the pain felt by the operators through a longitudinal study performed through NORDIC questionnaires. The results show that the results of ergonomic evaluation of operations can be used in a stochastic parametric model to assess the evolution of pains of the operators on long time scales.

Keywords: Vehicle dynamic systems, Autonomous vehicles
Abstract: This work investigates the problem of state estimation for nonlinear triangular systems having additional output measurements. The main objective of this paper is to propose observer design strategies that account for the available additional information, which is difficult to consider using the standard high-gain observer methodology. To deal with this challenge, we propose a novel observer design method to handle the additional output measurements. This method can be thought of as an extension of the standard high-gain observer by introducing a weighting matrix as a tuning parameter, which uses both the high-gain methodology and the LPV/LMI technique. The proposed approach was applied to a nonlinear longitudinal dynamic model that uses extra measurement information, in addition to the measured longitudinal position of the vehicle.

Keywords: Vehicle dynamic systems, Human and vehicle interaction, Design, control and monitoring of autonomous transportation systems
Abstract: This paper presents an unknown input observer (UIO)-based method to jointly estimate the vehicle dynamics and the driver torque within the framework of human-machine shared driving. To deal with the time-varying vehicle longitudinal speed, the vehicle dynamics is represented as a linear parameter-varying (LPV) model. Based on an unknown input (UI) decoupling technique, an LPV observer is designed, which can guarantee an asymptotic estimation performance of both the vehicle dynamics and the driver torque. Via Lyapunov stability theory, we propose sufficient conditions, expressed in terms of linear matrix inequalities, to design LPV UIO. High-fidelity Simulink-CarSim co-simulations are carried out to show the effectiveness of the proposed LPV UIO-based estimation method for driver-automation shared driving dynamics. Moreover, a comparative study is performed with a recent LPV estimation method to highlight the practical interests of the new solution.

Keywords: Neural networks, Vehicle dynamic systems
Abstract: This paper proposes a novel continuous-discrete (sampled data) time neural network (NSNN) observer for nonlinear systems. It can therefore be applied to systems with a high degree of non-linearity with no prior knowledge of the system dynamics. The proposed observer is a three-layer feedforward neural network that has been intensively trained using the error backpropagation learning algorithm, which includes an e-modification term to ensure robustness of the observer. A structure of the output predictor with a corrective term is added in the structure of the NN observer to overcome the problem of discrete time measurement. Simulations using MATLAB and CarSim are illustrated to demonstrate the performance of the proposed state observer strategy to reconstruct the state variables and parameters of a vehicle system.

Keywords: Modelling and control of road traffic networks, Human and vehicle interaction, Human factors in traffic and transportation control
Abstract: The vehicle nudging principle describes how a vehicle in the traffic flow induces a ‘pushing effect’ to its preceding vehicle. When combined with the traditional vehicle-following behaviour, the nudging principle enables bidirectional inter-vehicle interactions (look-ahead-and-behind). Unfortunately, despite numerical examples and traffic simulators indicating that nudging may improve the traffic flow, such results use artificially engineered scenarios. By using the NGSIM real-world traffic data sets, this work suggests that the nudging effect is a part of human-driven traffic as well, and can be suitably identified.

Keywords: Intelligent Transportation Systems, Sensor integration and perception, Multi-vehicle systems
Abstract: In this paper, we propose an integrated framework for safe intersection coordination of connected and automated vehicles (CAVs) in mixed traffic. An intelligent intersection is introduced as a central node to orchestrate state data sharing among connected agents and enable CAV to acknowledge the presence of human-driven vehicles (HDVs) beyond the line of sight of onboard sensors. Since state data shared between agents might be uncertain or delayed, we design the intelligent intersection to safely compensate for these uncertainties and delays using robust set estimation and forward reachability analysis. When the intersection receives state data from an agent, it first generates a zonotope to capture the possible measurement noise in the state estimate. Then, to compensate for communication and processing delays, it uses forward reachability analysis to enlarge the set to capture all the possible states the agent could have occupied throughout the delays. Finally, using the resulting set as the initial condition, a distributed model predictive control onboard the CAV will plan an invariant safe motion by considering the worst-case behavior of human drivers. As a result, the vehicle is guaranteed to be safe while driving through the intersection. A prototype of our proposed framework is implemented using the Small-Vehicles-for-Autonomy (SVEA) platform. The effectiveness of our framework is evaluated in experiments based on a challenging scenario where the collision would have occurred without efficient coordination.

Keywords: Navigation, guidance and control, Positioning systems, Autonomous vehicles
Abstract: Intelligent Transportation Systems (ITS) and Connected Autonomous Vehicles (CAV) are extremely active areas of research, which have attracted the attention of users, companies, and development agencies worldwide. The Global Navigation Satellite System (GNSS) is a main enabling technology, as ITS/CAV applications are fundamentally dependent on precise positioning systems. Among the recent GNSS technology advances that have allowed high accuracy to be achieved is Real-Time Precise Point Positioning (RT-PPP). Even though the Real-Time Service (RTS) of the International GNSS Service (IGS) provides various correction streams for RT-PPP deployment, some of them (e.g., those associated with ionospheric delays) still have limited accuracy. This has motivated local organizations and research institutes to deploy their own RT-PPP products, based on a denser regional network of GNSS Continuously Operating Reference Stations (CORS). The purpose of this work is to evaluate the performance of a recently established regional ionosphere RT-PPP product, from the Argentine University of La Plata (UNLP), which serves all of South America, with extensions to the Caribbean and Antarctica Peninsula. The main contributions of the work are: we show that RT-PPP using UNLP products outperforms RT-PPP using products exclusively from IGS-RTS; we benchmark positioning strategies, such as Post-Processed PPP (PP-PPP) and Relative Global Positioning System (RGPS); and, demonstrate the ability of Brazilian RT-PPP users employing Single-Frequency (SF) GPS code measurements and UNLP ionosphere products to achieve the CAV lane-level specification (i.e., 1-meter horizontal positioning accuracy at 95% probability). The article includes results from both static and moving experimental tests.

Keywords: Model predictive control of hybrid systems
Abstract: Due to the increasing use of nonlinear loads in modern power systems, harmonic currents have become a more prominent problem for power quality. Recently, a novel constrained linear state signal shaping model predictive controller has been proposed for shunt active power filter control. However, due to the high computational requirements of online quadratic programming solvers, this solution is not ready for real-time implementation. Therefore, the present work proposes the use of a linear state signal shaping explicit model predictive control formulation, such that the optimizations are done offline. To deal with the large memory footprint of the offline data generated, a tensor representation is introduced, which can then be compressed via tensor decomposition methods. The proposed approach was tested in simulation and was able to provide good results with a considerable reduction of the memory requirement.

Keywords: Estimation and filtering, Identifiability, Sensor networks
Abstract: Phasor Measurement Units (PMUs) are measurement devices that are installed on the buses of power network to measure voltage phasor data of a bus and current phasors of adjacent branches. With information on the power network impedance, a finite number of PMUs can be used to estimate the voltage on all buses, also known as the state of the power network. Traditionally, the problem of Optimal PMU Placement (OPP) focuses on having complete observability of the network state by strategically placing a minimum number of PMUs. A change in network topology caused, for example, by network switches, induces a change in the solution to the OPP problem. Since it is not feasible to change the locations of PMUs after every change in topology, the contribution of this paper is a PMU placement optimization algorithm which minimizes the variance on the estimated voltages and is also robust to the changes to the network topology. The robustness to the change in network topology is guaranteed by this optimization since it considers each change in the switch position as a different network topology and the optimization criterion is formulated as a multi-objective optimization function across all the different network topology caused by changing the switch positions. The proposed robust optimal placement algorithm can be implemented for both single current-channel PMU case and multi-current channel PMU case. The algorithm is tested on IEEE 14 bus network. The results show that the robust optimal placement can significantly reduce the effects of PMU measurement noise on the estimated states on average for different switch positions.

Keywords: Kalman Filtering, Estimation and filtering, Nonparametric methods
Abstract: Real-time monitoring of fuel injection is significant for improving the accuracy of the injection control in diesel engines. To obtain the real-time injection information, this paper proposes an optimal estimation method for the fuel injection rate of high-pressure common rail system using the Kalman filter algorithm. A dynamic mathematical model between the rail pressure variation and injection rate is established based on the mechanism of fuel flow process. Then the Kalman filter for injection optimal estimation is designed. Finally, the estimation performance is verified by simulation. The results show that the estimated injection rate converges fast, and the deviations of injection volume between the estimated results and the actual values are less than 5%. The designed Kalman filter has successfully produced the optimal estimation of the fuel injection process.

Keywords: Synthesis of stochastic systems, Stability and stabilization of hybrid systems, Stochastic hybrid systems
Abstract: The paper investigates the mean-square stability and stabilization issues for discrete- time semi-Markov jump linear systems with fragmentary probabilistic distributions of sojourn time. As the probability distributions of the sojourn time are difficult to be completely reconstructed, the obtained probability mass functions of the sojourn time from the current mode to the target mode can be fragmentary. In view of this phenomenon, we propose a method that makes full utilization of all the known probabilistic information of the sojourn time to establish less conservative criteria for the stability and stabilization of semi-Markov jump linear systems than the methods that only leverage completely known probability mass functions of the sojourn time.

Keywords: Estimation and filtering, Machine learning, Kalman filtering techniques in automotive control
Abstract: For state and parameter estimation in vehicles, Kalman filters, especially nonlinear extensions like the extended Kalman filter (EKF) and unscented Kalman filter (UKF), are very common. However, the estimation accuracy is highly dependent on the quality of the model used in the process update of the Kalman filter. Model errors can result from non-modeled dynamics that are either unknown or very difficult to describe. In recent years data-driven approaches for state estimation are the subject of research with promising results in estimation accuracy and reduced implementation effort. In this work, both a model-based method with an UKF and a data-driven approach based on recurrent neural networks (RNN) are implemented and combined to two hybrid methods for the application of state and parameter estimation in a truck-semitrailer for three different loading conditions. Hybrid estimation architectures promise to combine the advantages of model-based and data-driven methods to achieve better estimation accuracy than their standalone components. To the best knowledge of the authors, this work is the first to extend the field of hybrid state estimation to semitrailers estimating the truck steering angle, articulation angle, and the trailer's lateral and vertical tire forces. Four estimation architectures (an UKF, one purely data-driven method, and two hybrid methods) are optimized and compared to each other regarding estimation accuracy. The UKF is optimized with a particle swarm optimization (PSO) while the hyperparameters of the data-driven method are tuned with the asynchronous successive halving algorithm (ASHA) to result in a fair comparison. All methods are developed and compared based on an experimental data set from a test vehicle.

Keywords: Predictive control, Randomized algorithms
Abstract: This paper studies a sample-based model predictive control using Monte Carlo sampling method. Different from gradient-based MPC, MCMPC is able to introduce discontinuous phenomena such as collisions into the system which makes it suitable for robot explorations in narrow space. In this study, MCMPC is utilized for the control of an underwater snake robot. Control of each joint angle as inputs may however highly increase the dimension of samplings and prediction steps, therefore we combine curvature derivative control (CDC) with MCMPC so that only one control input is needed to be generated for controlling whole snake robot locomotion without considering the quantity of links. The proposed MCMPC algorithm generates variable gait pattern while minimizing the difference between the position of center of mass of snake robot and the target point in the designed cost function. Experiments for snake robot consists of 5,10 and 20 links are tested, depends on the trajectory results we show that by introducing CDC into MCMPC, smoother locomotion could be achieved. Underwater snake robot can reach the set point and change its locomotion during operations with single control input by utilizing the proposed control method.

Keywords: Autonomous underwater vehicles, Control architectures in marine systems, Teleoperation
Abstract: This paper focuses on the description of the Blue-RoSES project, an European funded activity devoted to the development and exploitation of a reliable infrastructure for the remote employment of Remotely Operated Vehicles (ROVs) for both professional and recreational tasks. The paper firstly provides an overall view of the concept and then summarize the main modules composing the architecture and responsible for a reliable and stable remote piloting framework for ROVs. Analysis of the critical issues related to the remote piloting and control is carried out and measurements of the communication lags and delays are performed to obtain a statistical characterization, also providing a quantitative measure of the communication service. The reliability and performance of the complete system are demonstrated through experimental campaigns in real case scenarios.

Keywords: Marine system navigation, guidance and control, Autonomous surface vehicles, Dynamic positioning
Abstract: We propose a new geometric algorithm for path planning of maritime vehicles, which partially meets the Convention on the International Regulations (COLREGs) for preventing collisions at sea. To carry out risk assessment with obstacles, one of the decision variables used is the intersection of the Line-of-Sight (LOS) with the obstacle's position. For dynamic obstacles, that position is based on the projection of the obstacle moving forward in time assuming a constant speed and heading. Following a positive risk of collision, a (starboard or portside) manoeuvre is applied. The algorithm is designed to adhere to COLREGs steering rules automatically, in particular, rules 13-16. Unlike using a binary decision variable for risk assessment, which is a common method of choice in the literature, the proposed approach utilises a fuzzy risk index to determine the level of collision risk, minimising the occurrence of false positives. The computational complexity of the proposed algorithm is relatively low making it appealing for real-time implementation. We report encouraging results via simulation analysis for a range of scenarios involving both static and dynamic obstacles.

Keywords: Autonomous surface vehicles, Trajectory and path planning, Intelligent Transportation Systems
Abstract: This paper provides a collision avoidance perspective to maritime autonomy, in the shift towards Maritime Autonomous Surface Ships (MASS). In particular, the paper presents the developments related to the Greenhoper, Denmark's first autonomous harbour bus. The collision and grounding avoidance scheme, called the Short Horizon Planner (SHP), is described and discussed in detail. Furthermore, the required autonomy stack for facilitating safe and rule-compliant collision avoidance is presented. The inherent difficulties related to adhering to the COLREGs are outlined, highlighting some of the operational constraints and challenges within the space of autonomous ferries and harbour buses. Finally, collision and grounding avoidance is demonstrated using a simulation of the whole Greenhopper autonomy stack.

Keywords: Neural networks, Autonomous underwater vehicles, Sensing
Abstract: Recent developments in marine technologies allow underwater vehicles to perform survey missions for data collection in an automatic way. The scientific community is now focusing on endowing these vehicles with strong perception capabilities, aiming at full autonomy and decision-making skills. Such abilities would bring benefits to a wide range of field applications, e.g. Inspection and Maintenance (I&M) of man-made structures, port security, and marine rescue. Indeed, most of these tasks are currently carried out employing remotely operated vehicles, making the presence of humans in water necessary. Projects like Metrological Evaluation and Testing of Robots in International CompetitionS (METRICS), funded by the European Commission, are promoting research on this field by organising events such as the Robotics for Asset Maintenance and Inspection (RAMI) competition. In particular, this competition requires participants to develop perception techniques capable of identifying a set of specific targets. Within such context, this paper presents an algorithm able to detect and classify Objects of Potential Interest (OPIs) in underwater camera images. First, the proposed solution compensates for the quality degradation of underwater images by applying color enhancement and restoration procedures. Then, it exploits deep-learning techniques, as well as color and shape based methods, to recognize and correctly label the predefined OPIs. Preliminary results of the implemented neural network using restored images are provided, and a mean Average Precision (mAP) of about 92% was achieved on the dataset provided to the RAMI competition participating teams by the NATO Science and Technology Organization Centre for Maritime Research and Experimentation (STO CMRE).

Keywords: Nonlinear and optimal marine system control, Autonomous underwater vehicles, Unmanned marine vehicles
Abstract: In this paper we explore the optimal positioning of an underwater snake robot (USR) for energy harvesting in the wake of a bluff body. The USR and fluid are simulated jointly using a coupled vortex particle-mesh method and multi-body system solver. The power dissipated in the damped joints of the robot is used as a proxy for the harvested energy. Furthermore, the effect of different damper coefficients on power dissipation is explored. An extremum-seeking control (ESC) scheme for nonlinear systems with time-varying steady-state solutions is employed to optimize the horizontal placement of the robot. It is found that the dissipated power does have a clear optimum. Moreover, the horizontal position of the USR under the ESC scheme is demonstrated to converge to a vicinity of the optimal position.

Keywords: Acoustic-based networked control and navigation, Marine system navigation, guidance and control, Unmanned marine vehicles
Abstract: Cooperation among heterogeneous marine vehicles can offer several benefits to underwater applications, such as supporting the navigation of submerged robots by leveraging the information available to a surface vehicle. Within such context, this paper describes an acoustic positioning and communication protocol developed for the localisation of multiple Autonomous Underwater Vehicles (AUVs) by means of an Autonomous Surface Vehicle (ASV) equipped with an Ultra-Short BaseLine (USBL) device. To avoid packet collisions, access to the acoustic channel is managed through a time slot division mechanism. Unlike a classical Time Division Multiple Access (TDMA) protocol, the approach considered in this work consists of a centralised scheduling handled by the ASV, in which AUVs are localised at regular time epoch but can transmit only when queried. This allows to prioritise the number of position measurements taken by the ASV and then relayed back to the submerged vehicles, with the final goal of improving the navigation accuracy of the latter. The solution is also designed to handle the latency of acoustic communication and to correctly associate each position measurement with the corresponding acquisition time. Furthermore, it does not require any a priori synchronisation between the clocks of the vehicles involved and ensures complete decoupling between their navigation algorithms. Experimental activities in very-shallow waters, involving an ASV and two target nodes, were carried out to validate the multi-AUVs positioning system and to provide a characterisation of its performance.

Keywords: Data-driven optimal control, Predictive control
Abstract: This paper proposes a method to encourage safety in Model Predictive Control (MPC)-based Reinforcement Learning (RL) via Gaussian Process (GP) regression. The framework consists of 1) a parametric MPC scheme that is employed as model-based controller with approximate knowledge on the real system’s dynamics, 2) an episodic RL algorithm tasked with adjusting the MPC parametrization in order to increase its performance, and 3) GP regressors used to estimate, directly from data, constraints on the MPC parameters capable of predicting, up to some probability, whether the parametrization is likely to yield a safe or unsafe policy. These constraints are then enforced onto the RL updates in an effort to enhance the learning method with a probabilistic safety mechanism. Compared to other recent publications combining safe RL with MPC, our method does not require further assumptions on, e.g., the prediction model in order to retain computational tractability. We illustrate the results of our method in a numerical example on the control of a quadrotor drone in a safety-critical environment.

Keywords: Predictive control, Data-driven optimal control
Abstract: Model predictive control is a powerful advanced control technique to deal with complex nonlinear systems with constraints. Despite recent advances in computing hardware and optimization algorithms, solving the nonconvex optimization problems associated to nonlinear model predictive control in real time can still be intractable.

There are several possibilities to alleviate the challenge of computational complexity. An increasingly popular approach is to use neural networks to approximate the mapping between current system state and the corresponding optimal control input, which is implicitly defined by an optimization problem. These approaches typically need a large amount of data. Although the data can be generated offline by sampling the state space and solving many model predictive control problems in advance, it can become intractable especially for increasing system dimensions. In this work, we propose to use the parametric sensitivities of the nonlinear optimization problems as additional information that is provided to the neural network during training. In this way, the neural network not only fits the solution of the optimization problems itself but also its gradients. This training method, usually called Sobolev training, can lead to a higher accuracy in the approximation for the same amount of data. We illustrate the potential of the approach with a simulation study of a nonlinear chemical reactor.

Keywords: Data-driven optimal control, Predictive control, Data-driven robust control
Abstract: Recent publications have laid a solid theoretical foundation for the combination of Reinforcement Learning and Model Predictive Control, in view of obtaining high-performance data-driven MPC policies. Early practical results, both in simulation and in experiments, have shown the potential of this combination but have also revealed certain challenges. In addition, the technical complexity of these results makes it difficult for interested readers to gather the fundamental ideas and principles behind this combination. This paper aims to provide a coherent and more accessible picture of these results and to offer significantly deeper and more mature insights into their meaning than has been proposed before. It also aims at identifying the current challenges in the field.

Keywords: Predictive control, Data-driven optimal control, Nonlinear predictive control
Abstract: There are different approaches to tune a control policy that would result in a desired closed-loop performance. The typical design flow involves tuning the control policies offline using (high-fidelity) simulators until satisfactory performance is achieved. This paper on the other hand, considers the problem of tuning parameterized control policies directly by interacting with the real system that also has safety-critical constraints. We use safe Bayesian optimization using interior-point methods to tune the parameters online that guarantees constraint satisfaction with high probability. The proposed framework is applied to a personalized artificial pancreas system for type 1 diabetes. The paper shows that the parameterized control policy used for blood glucose regulation can be safely tuned to personalize the controller for each individual patient using our approach and thus improve its performance.

Keywords: Data-based control, Parametric optimization
Abstract: In mobile robot applications, the trajectory tracking task hides several difficulties, including the choice of the setpoint and the search for an acceptable trade-off between performance and computational constraints. In this work, we discuss practical issues of a Reinforcement Learning (RL) based Model Predictive Control (MPC) tuning approach by focusing on a specific mobile robot application, where the objective is to maximize the velocity, while keeping the robot within the track bounds. Among others, we show that softening the latter constraints allows us to obtain a RL-tuned tracking controller with the same performance of an economic nonlinear MPC formulation, but requiring significantly less computational resources.

Keywords: Nonlinear predictive control, Parametric optimization, Data-driven optimal control
Abstract: This study deals with the control methods for automation of the loading operation of hydraulic excavators for mass production. In the loading operation, it is necessary to move the bucket to the target position without spilling soil from the bucket. Since the loading operation is a combined operation of the front equipment and the swing operation, it is not easy to generate the ideal trajectory. We apply Model Predictive Control (MPC) to this problem and investigate a method to realize ideal loading operation by giving only the coordinates of the loading position. When using MPC, the challenge is that a lot of worker-hours are spent on tuning the weights of the objective function. We also propose an efficient weight tuning method to solve this problem, with a view to adapting it to hydraulic excavators for mass production. We have confirmed the effectiveness of the proposed method through numerical simulations.

Keywords: Robustness analysis, Convex optimization, Uncertain systems
Abstract: We propose novel time-domain dynamic integral quadratic constraints with a terminal cost for exponentially weighted slope-restricted gradients of not necessarily convex functions. This extends recent results for subdifferentials of convex function and their link to so-called O'Shea-Zames-Falb multipliers. The benefit of merging time-domain and frequency-domain techniques is demonstrated for linear saturated systems.

Keywords: Robustness analysis, System analysis and optimization
Abstract: We recapitulate the notion of phase change rate maximization and demonstrate the usefulness of its solution on analyzing the robust instability of a cyclic network of multi-agent systems subject to a homogenous multiplicative perturbation. Subsequently, we apply the phase change rate maximization result to two practical applications. The first is a magnetic levitation system, while the second is a repressilator with time-delay in synthetic biology.

Keywords: Robustness analysis, Stability of nonlinear systems, Robust linear matrix inequalities
Abstract: This paper considers the stability analysis of a class of sampled data Lur'e systems with possibly aperiodic sampling times. The analysis conditions are formulated in an IQC framework where i) the nonlinear component, as usual, is modelled using sector and slope restrictions which have well-known integral quadratic constraints associated with them; and ii) the sample-and-hold operator is represented, after loop transformations as an operator which satisfies certain other integral quadratic constraints. The arising analysis conditions can then be expressed as a frequency domain condition and a time domain constraint. Under various assumptions these can be transformed into linear matrix inequality conditions. Quite logically, when the bound on the sample interval tends to zero, the continuous time stability conditions for Lur'e systems are recovered.

Keywords: Robustness analysis, Data-driven robust control, Robust linear matrix inequalities
Abstract: Starting from a linear fractional representation of a linear system affected by constant parametric uncertainties, we demonstrate how to enhance standard robust analysis tests by taking available (noisy) input-output data of the uncertain system into account. Our approach relies on lifting the system and the construction of data-dependent multipliers. It leads to a test in terms of linear matrix inequalities which guarantees stability and performance for all systems compatible with the observed data if it is in the affirmative. In contrast to many other data-based approaches, prior physical or structural knowledge about the system can be incorporated at the outset by exploiting the power of linear fractional representations.

Keywords: Robustness analysis, Uncertain systems, Convex optimization
Abstract: This note revisits extended Linear Matrix Inequality (LMI) formulation of performance analysis, which incorporates static multiplier, for parametrically multi-affine discrete-time Linear Time-Invariant Parameter-Dependent (LTIPD) and Linear Periodically Time-Varying (LPTV) systems using lifting technique. Almost the same analysis conditions for parametrically affine discrete-time LTIPD/LPTV systems as the ones shown in this note have already been proposed in the literature, we thus focus on the simple derivations of the conditions using only Elimination lemma. To this end, a variant of extended LMI formulation, whose dimensional sizes are slightly larger than the conventional formulation with additional auxiliary matrices, is proposed in this note. It is also shown that our derived variant directly leads to a parametrically multi-affine extended LMI formulation for our addressed problem.

Keywords: Robustness analysis, Time-varying systems, Uncertain systems
Abstract: The paper presents a novel approach for robustness analysis of nonlinear dynamic systems in the vicinity of a reference trajectory. The approach linearizes the system with respect to a nominal trajectory and calculates a guaranteed upper bound on the worst-case gain. In contrast to existing methods rooted in linear time-varying systems analysis, the approach accurately includes perturbations that drive the system away from the reference trajectory. The approach further includes a bound for the error associated with the time-varying linearization. Hence, the results obtained in the linear framework provide a valid upper bound for the worst-case performance of the nonlinear system. The calculation of the upper bound relies on the dissipation inequalities formulated in the framework of integral quadratic constraints. It is therefore computationally much cheaper than sample-based methods such as Monte Carlo simulation. The feasibility of the approach is demonstrated on a numerical example.

Keywords: Nonlinear system identification, Frequency domain identification, Stochastic system identification in signal processing
Abstract: Identification of bilinear systems subject to white inputs is studied. Assuming bilinearity we use the crosscovariances between the output and the higher order Hermite polynomials of the input for estimating the coefficients of the Hermite series expansion of the output. The parameters of the bilinear model are obtained via a balanced factorization of the appropriately constructed Hankel matrix. The skewness of the singular values of the estimated Hankel matrix is applied for testing the order selection for the bilinear model.

Keywords: Machine learning, Nonlinear system identification, Autonomous vehicles
Abstract: The uncertainty in the prediction calculated using the delta method for an overparameterized (parametric) black-box model is shown to be larger or equal to the uncertainty in the prediction of a canonical (minimal) model. Equality holds if the additional parameters of the overparameterized model do not add flexibility to the model. As a conclusion, for an overparameterized black-box model, the calculated uncertainty in the prediction by the delta method is not underestimated. The results are shown analytically and are validated in a simulation experiment where the relationship between the normalized traction force and the wheel slip of a car is modelled using e.g., a neural network.

Keywords: Nonlinear system identification, Machine learning, Modeling and control of agriculture
Abstract: In this paper, the functioning and performance of a recently proposed weighting approach for dynamic datasets is demonstrated for nonlinear system identification. The overall goal of the new method is to improve data-driven model performance. This is achieved by increasing the weight of training data points in areas of the regressor space that are underrepresented during training, while decreasing the weight for overrepresented data points. The method is based on estimating the parameters of a model using weighted error norms, where the weighting is carried out with the inverse values of the kernel density estimated in the regressor space. The performance of the new method is investigated using nonlinear ARX models for the identification of (i) an artificial Hammerstein process and (ii) a nonlinear real-world process. Overall, the new weighting method shows significantly improved performance on the simulation error. It is demonstrated that performance is improved most notably in areas that are underrepresented in the training data.

Keywords: Nonlinear system identification, Machine learning in modelling, prediction, control and automation, Time series modelling
Abstract: Despite the immense success of neural networks in modeling system dynamics from data, they often remain physics-agnostic black boxes. In the particular case of physical systems, they might consequently make physically inconsistent predictions, which makes them unreliable in practice. In this paper, we leverage the framework of Irreversible port-Hamiltonian Systems (IPHS), which can describe most multi-physics systems, and rely on Neural Ordinary Differential Equations (NODEs) to learn their parameters from data. Since IPHS models are consistent with the first and second principles of thermodynamics by design, so are the proposed Physically Consistent NODEs (PC-NODEs). Furthermore, the NODE training procedure allows us to seamlessly incorporate prior knowledge of the system properties in the learned dynamics. We demonstrate the effectiveness of the proposed method by learning the thermodynamics of a building from the real-world measurements and the dynamics of a simulated gas-piston system. Thanks to the modularity and flexibility of the IPHS framework, PC-NODEs can be extended to learn physically consistent models of multi-physics distributed systems

Keywords: Grey box modelling, Nonlinear system identification, Continuous time system estimation
Abstract: This study presents a set-based method to estimate unknown parameters of continuous-time structured nonlinear systems under uncertainties. Sets containing the parameters are found by using data sets and structures of the systems. Such estimation is challenging because the uncertainties are caused by disturbances, data noises, and partially unknown dynamics. In addition, the estimation for continuous-time nonlinear systems faces nonlinear problems that are difficult to address. To overcome these issues, the proposed set-based method solves linear programming (LP) problems iteratively instead of the nonlinear problems. The inequalities in the LP are derived such that they are consistent with the true parameters, using the boundedness of the uncertainties. A theoretical analysis guarantees that solving the iterative LP gives sets containing the true parameters.

Keywords: Machine learning, Nonlinear system identification, Estimation theory
Abstract: A regression model with more parameters than data points in the training data is overparametrized and has the capability to interpolate the training data. Based on the classical bias-variance tradeoff expressions, it is commonly assumed that models which interpolate noisy training data are poor to generalize. In some cases, this is not true. The best models obtained are overparametrized and the testing error exhibits the double descent behavior as the model order increases. In this contribution, we provide some analysis to explain the double descent phenomenon, first reported in the machine learning literature. We focus on interpolating models derived from the minimum norm solution to the classical least-squares problem and also briefly discuss model fitting using ridge regression. We derive a result based on the behavior of the smallest singular value of the regression matrix that explains the peak location and the double descent shape of the testing error as a function of model order.

Keywords: Machine learning, Closed loop identification
Abstract: Quantum detector tomography is a fundamental technique for calibrating quantum devices and thus lay foundations for quantum information processing tasks. In this work, we propose a quantum detector tomography method that employs deep neural networks to reconstruct quantum detectors from a set of probe states with high efficiency. Numerical results demonstrate that the proposed method exhibits a significant potential to estimate phase-insensitive detectors.

Keywords: Closed loop identification, Continuous time system estimation
Abstract: Quantum state tomography and quantum detector tomography are two main problems in quantum system identification. In this paper, we study state tomography and detector tomography with prior knowledge about the true rank of the unknown state/detector. In this case, we propose a sufficient condition for non-adaptive tomography algorithms to achieve the optimal infidelity scaling O(1/N) on the number of state copies N . Then we present two examples of such non-adaptive quantum state tomography and quantum detector tomography algorithms. Numerical examples demonstrate the effectiveness of our algorithms.

Keywords: Machine learning
Abstract: It is challenging to identify the state of many-body quantum systems, as recovering density matrices underlying the quantum state typically requires computational resources scale exponentially to the system size. In this work, we introduce the variational diffusion model (VDM) to efficiently learn high-dimensional quantum distributions with high fidelity, which is essential to realize the fast reconstruction of quantum states. We build up the VDM suitable for dealing with the high-dimensional quantum samples, and then perform numerical experiments to test our model and other autoregressive models, including recurrent neural network and transformer. It is found that the VDM can achieve a modest better performance with fewest parameters than other two to learn the distribution as desired. Our results pave the way to applying diffusion models to solve hard problems in the quantum domain.

Keywords: Fault detection and diagnosis, Control of quantum and Schroedinger systems
Abstract: Quantum technologies have been considered as a method to enhance cyber-security due to the unique properties of quantum systems. However, quantum systems themselves could also suffer from strategic attacks. In this paper, we formulate the relationship between the strategic attacker and a passive quantum detector into a Stackelberg game. By deriving the optimal detection rule at equilibrium, we study the detection performance of the detector under the setting of finite sample size of corrupted observations. Under mild assumptions, we show that the miss. We illustrate the general decaying results of the miss rate numerically, depicting that the passive detector manages to achieve a miss rate and a false alarm rate both exponentially decaying to zero given infinitely many quantum states, although at a much slower rate than a quantum non-adversarial counterpart. Finally, we adopt our formulations upon a case study of detection with quantum radars.

Keywords: Large scale optimization problems, Control of quantum and Schroedinger systems, Control of heat and mass transfer systems
Abstract: In this work, a discrete heat distribution (DHD) problem is formulated and solved using the quantum approximate optimization algorithm (QAOA). The goal of the DHD problem is to suggest a reasonable valve configuration that finds a compromise between room temperature comfort and desired water flow. The DHD problem is formulated as an integer quadratic programming (IQP) problem, which is an NP-hard combinatorial optimization problem. For a large number of decision variables, this problem becomes computationally intractable to be solved for an exact solution and rather an approximate solution is favorable, making QAOA a potential candidate to solve this problem. For the DHD problem to be solved using the QAOA, it is converted first to a quadratic unconstrained binary optimization (QUBO) problem. The algorithm is implemented and simulated for various circuit depths and tested on IBM quantum computer. The results show that the algorithm can solve the problem for low circuit depths. This paper aims at showing a use case of NISQ devices in control engineering and suggests new applications of the QAOA in the design of controllers for dynamical systems.

Keywords: Stochastic adaptive control, Stochastic control, Control of quantum and Schroedinger systems
Abstract: Fast learning of uncertain parameters is crucial for control performance of adaptive control. This paper proposes two optimization problems for stochastic approximation and discusses a faster convergence rate of adaptive control for a class of stochastic systems based on the optimal solution. Under a specific condition, we derive the optimal adaptive gain explicitly and discuss how faster convergence of a state for an adaptive quantum control problem achieves based on the optimal adaptive gain.

Keywords: Distributed optimization for large-scale systems, Quantized control, Distributed control and estimation
Abstract: In this paper, we consider the unconstrained distributed optimization problem, in which the exchange of information in the network is captured by a directed graph topology, thus, nodes can only communicate with their neighbors. Additionally, in our problem, the communication channels among the nodes have limited bandwidth. In order to alleviate this limitation, quantized messages should be exchanged among the nodes. For solving this distributed optimization problem, we combine a gradient descent method with a distributed quantized consensus algorithm (which requires the nodes to exchange quantized messages and converges in a finite number of steps). Specifically, at every optimization step, each node (i) performs a gradient descent step (i.e., subtracts the scaled gradient from its current estimate), and (ii) performs a finite-time calculation of the quantized average of every node’s estimate in the network. As a consequence, this algorithm approximately mimics the centralized gradient descent algorithm. We show that our algorithm asymptotically converges to a neighborhood of the optimal solution with linear convergence rate. The performance of the proposed algorithm is demonstrated via simple illustrative examples.

Keywords: Distributed optimization for large-scale systems, Security in networked control systems, Distributed control and estimation
Abstract: This paper presents a continuous-time resilient distributed optimization algorithm based on competitive interaction design method on connected graphs in the presence of adversaries. The competitive interaction method allows us to design a network that protects the multi-agent systems from adversaries without requiring high network connectivity. In addition, the proposed algorithm does not require the global information about the number of adversaries. First, we show that the proposed distributed algorithm solves the resilient distributed optimization problem with no attack on the communication links. Second, we show that the proposed continuous-time distributed optimization algorithm on connected graphs converges to the small neighborhood of the optimal solution in the presence of cyber-attacks onto the communication channel. Simulations are presented to illustrate our theoretical results.

Keywords: Distributed optimization for large-scale systems, Game theories, Graph-based methods for networked systems
Abstract: The vertex cover (VC) problem has been studied for past few years, and some centralized and distributed algorithms have been proposed. In this paper, we propose the so-called random memory length adaptive (RMLA) algorithm for the VC problem in networks. Any initial state can reach the stable pure strict Nash equilibrium state through finite iterations by the RMLA algorithm. We find that by setting the minimum edge-keeping probability reasonably, the convergence rapidity and accuracy can be improved simultaneously. Our algorithm also removes the requirement for consistent node memory length. Through theoretical analysis and intensive numerical simulations, we verify the convergence and effectiveness of the RMLA algorithm.

Keywords: Distributed optimization for large-scale systems, Large scale optimization problems, Smart grids
Abstract: We address the optimal operation of a large-scale multi-agent system where agents have to set their own continuous and/or discrete decision variables so as to jointly minimize the sum of local linear performance indices while satisfying local and global linear constraints. When the number of discrete decision variables is large, solving the resulting Mixed Integer Linear Program becomes computationally demanding, and often impossible in practice. Inspired by some recent methods in the literature, we propose a decentralized iterative scheme that recovers computational tractability by decomposing the dual of the MILP problem into lower-dimensional MILPs, one per agent, and obtains feasibility of the recovered primal solution by introducing a fictitious tightening of the global constraints. The tightening is updated in an adaptive fashion according to an heuristic strategy which allows it to both increase and decrease throughout the iterations, depending on the mismatch between the recovered mixed-integer primal solution and the solution to the relaxed linear problem associated with the current tightening. The procedure is shown to be effective and to outperform state-of-the-art alternative resolution schemes in a benchmark example on optimal charging of a fleet of electric vehicles.

Keywords: Distributed optimization for large-scale systems, Consensus, Multi-agent systems
Abstract: In this study, a Newton consensus method is proposed for the distributed optimization of a multi-agent systems operating over strongly connected digraphs. The approach proposes an approximate Newton step for both the primal and dual problems that can be implemented in a completely decentralized fashion. The asymmetry in the communication network is addressed by computing an approximate Newton step that only requires the out-Laplacian. The proposed Newton consensus approach does not require the exchange of derivative information between agents. In addition, the optimization approach avoids the explicit inversion of the approximate Hessian information for the computation of the Newton step. A simulation study demonstrates the effectiveness of the technique.

Keywords: Distributed optimization for large-scale systems, Multi-agent systems
Abstract: We solve large-scale mixed-integer linear programs (MILPs) via distributed asynchronous saddle point computation. To solve a MILP, we relax it with a nonlinear program approximation whose accuracy tightens as the number of agents increases relative to the number of coupled constraints. Next, we form an equivalent Lagrangian saddle point problem, and create a regularized Lagrangian that is strongly-convex-strongly-concave. We then develop a parallelized algorithm to compute saddle points of the regularized Lagrangian. This algorithm partitions problems into blocks and it is shown to tolerate asynchrony in the computations and communications of primal and dual variables. Suboptimality bounds and convergence rates are presented for convergence to a saddle point. The suboptimality bound includes (i) the regularization error induced by regularizing the Lagrangian and (ii) the suboptimality gap between solutions to the original MILP and its relaxed form. Simulation results illustrate these theoretical developments in practice, and show that relaxation and regularization together have only a mild impact on the quality of solution obtained.

Keywords: Estimation and filtering, Identification for control, Mechanical and aerospace estimation
Abstract: This paper revisits the previously proposed linear asymptotic observer of the motion state variables with nonlinear friction and provides a robust design suitable for both, transient presliding and steady-state sliding phases of the relative motion. The class of motion systems with the only measurable output displacement is considered. The reduced-order Luenberger-type observer is designed based on the obtained simplified state-space representation with a time-varying system matrix. The resulted observation error dynamics proves to be robust and appropriate for all variations of the system matrix, which are due to the nonlinear spatially-varying friction. A specially designed tribological setup to accurately monitor the relative motion between two contacting friction surfaces is used to collect the experimental data of the deceleration trajectories when excited by a series of impulses. The performance of the state estimation using the proposed observer is shown based on the collected experimental data.

Keywords: Estimation and filtering, Estimation theory, Bayesian methods
Abstract: This paper deals with state estimation of stochastic models with linear state dynamics, continuous or discrete in time. The emphasis is laid on a numerical solution to the state prediction by the time-update step of the grid-point-based point-mass filter (PMF), which is the most computationally demanding part of the PMF algorithm. A novel way of manipulating the grid, leading to the time-update in form of a convolution, is proposed. This reduces the PMF time complexity from quadratic to log-linear with respect to the number of grid points. Furthermore, the number of unique transition probability values is greatly reduced causing a significant reduction of the data storage needed. The proposed PMF prediction step is verified in a numerical study.

Keywords: Estimation and filtering, Stochastic system identification, Continuous time system estimation
Abstract: Estimation of states in stochastic differential equations with state dependent diffusion is known to be difficult. Previous research recommend the higher order extended Kalman filter or the Lamperti transform method for this case. This paper shows that a new developed method, based on the unscented Kalman filter, is superior for two simulated stochastic differential equation systems.

Keywords: Estimation and filtering, Kalman Filtering, Estimation theory
Abstract: An easy-to-implement method for nonlinear state estimation for ill-conditioned systems is proposed. By propagating standard deviations and correlations instead of the covariance in the unscented Kalman filter (UKF), the condition numbers of relevant matrices are reduced. The reduction in the condition number is related to the scaling of the problem. Hence, what we propose is a normalization method that acts as an “auto-scaler”. Compared to other methods in state estimation for ill-conditioned systems, our proposed method factors the covariance matrix into physically meaningful statistics which can be used to check for filter divergence online. The method is compared to a standard UKF in a case study and shows a significant reduction in the condition number.

Keywords: Estimation and filtering, Kalman Filtering, Stochastic system identification
Abstract: We consider the problem of joint parameter estimation and smoothing in structured linear systems using the expectation maximization (EM) framework. Specifically, we explore how partially known sparsity structures in the estimation model can be leveraged to improve the computation speed and performance of the considered EM approaches. We use these ideas to generalize a recently proposed GraphEM algorithm to a linear time-varying setting, where the sparsity structures may vary in time. We obtain a biconvex form of the majorizing function in the M-step, which is minimized subject to an l1-regularization using a Douglas-Rachford proximal splitting algorithm. Numerical results using a satellite positioning example shows significant improvements in the estimation errors and an F1-score that quantifies model sparsity.

Keywords: Estimation and filtering, Linear systems, Infrastructure (including energy, telecoms, political, physical, etc.)
Abstract: To generate accurate time scales, such as national standard time in each country and network time in communication systems, a novel prediction algorithm is proposed for the atomic clock ensembles in which the system is unobservable. In the proposed prediction algorithm, to deal with unobservable state space, we consider a steady-state Kalman filter and an intuitive prediction algorithm to separately treat the observable and unobservable state spaces after Kalman canonical decomposition. It turns out that the basis selection plays an important role in affecting prediction performance. We compare the performance of the proposed algorithm with the conventional one and discuss the possibility of prediction improvements by altering the transformation matrix. A couple of numerical examples are presented to illustrate the efficacy of our results.

Keywords: Event-triggered and self-triggered control, Digital implementation, Nonlinear time-delay systems
Abstract: In this paper, the stabilization problem of nonlinear time-delay systems via quantized sampled-data event-based controllers is investigated. Fully nonlinear (i.e., possibly non-affine in the control) systems affected by state delays are studied. Sufficient conditions are provided for the existence of a suitably fast sampling and of an accurate quantization of the input/output channels such that the digital implementation of the continuous-time controller at hand, updated through a proposed event-triggered mechanism, ensures the semi--global practical stability property, with arbitrarily small final target ball of the origin, of the related closed-loop system. A spline approximation methodology is used in order to cope with the problem of the possible non-availability in the buffer of suitable past values of the system variables needed for the digital implementation of the controller. The stabilization in the sample-and-hold sense theory is used as a tool to prove the results. In the theory here developed, the case of non-uniform quantization of the input/output channels and the case of aperiodic sampling are both included. The proposed theoretical results are validated through an application concerning the plasma glucose regulation problem in type-2 diabetic patients.

Keywords: Event-triggered and self-triggered control, Multi-agent systems, Control over networks
Abstract: Event-triggered control has the potential to provide a similar performance level as time-triggered (periodic) control while triggering events less frequently. It therefore appears intuitive that it is also a viable approach for distributed systems to save scarce shared network resources used for inter-agent communication. While this motivation is commonly used also for multi-agent systems, a theoretical analysis of the impact of network effects on the performance of event- and time-triggered control for such distributed systems is currently missing. With this paper, we contrast event- and time-triggered control performance for a single-integrator consensus problem under consideration of a shared communication medium. We therefore incorporate transmission delays and packet loss in our analysis and compare the triggering scheme performance under two simple medium access control protocols. We find that network effects can degrade the performance of event-triggered control beyond the performance level of time-triggered control for the same average triggering rate if the network is used intensively. Moreover, the performance advantage of event-triggered control shrinks with an increasing number of agents and is even lost for sufficiently large networks in the considered setup.

Keywords: Event-triggered and self-triggered control, Multi-agent systems, Distributed control and estimation
Abstract: This paper deals with a cloud-mediated self-triggered state synchronization control method of a linear multi-agent system subject to disturbances. In the cloud-mediated self-triggered control approach, each agent asynchronously accesses the cloud repository to get past information on its neighboring agents. Then, the agent predicts the future behavior of its neighbors as well as of its own, and locally determines its next access time to the cloud repository. In this paper, we propose a cloud-mediated state synchronizing controller design method in the presence of disturbances. The uncertainty caused by the disturbances is appropriately handled in the self-triggering rule. It is also proved that the closed-loop system does not exhibit any Zeno behaviors.

Keywords: Event-triggered and self-triggered control, Networked robotic systems, Consensus
Abstract: Consensus and synchronization are fundamental concepts for the coordination of cooperating multi-robot teams. Applications like cooperative manipulation may require not only synchronization of the position, but also of the orientation of the individual agents. The pose of the agent can be described within the special Euclidean group and common results for coordination have to be adapted. We propose a control framework for full-pose synchronization in SE(3) for a team of Euler-Lagrange agents, relying only on relative information in the absence of a global coordinate frame. The framework consists of an inner loop for feedback linearization and an outer loop for pose synchronization. The measurements are taken by external sensors and communicated to the respective agents via a common communication network. To deal with limited communication bandwidth, we propose an event-triggered update strategy for the relative measurements. Finally, the efficacy of the proposed control and triggering law is illustrated in simulations.

Keywords: Sensor networks, Consensus, Distributed control and estimation
Abstract: Distributed sensor networks have gained interest thanks to the developments in processing power and communications. Event-triggering mechanisms can be useful in reducing communication between the nodes of the network, while still ensuring an adequate behaviour of the system. However, very little attention has been given to continuous-time systems in this context. In this work, we propose a strategy for distributed state estimation in sensor networks, based on average dynamic consensus of the continuous measurements. While communication between nodes is discrete and heavily reduced due to the event-triggering mechanism, our method ensures that the nodes are still able to produce a continuous estimate of the global average measurement and the state of the plant, within some tuneable error bounds.

Keywords: Fault detection and diagnosis, Nonparametric methods, Intelligent Transportation Systems
Abstract: Predictive Maintenance (PdM) aims to estimate the optimal moment when the maintenance of an industrial asset should be performed according to its actual health status. The goal is to minimise the costs, by finding the optimal point where the sum of the prevention cost and the repair cost is at the lowest. Data-driven solutions can help build more cost-efficient maintenance strategies by providing a tool to predict whether the asset is close to suffering a real breakdown. This paper focuses on Survival-Analysis-based Predictive Maintenance applied to the case of Battery Electric Trucks (BET). Cox Proportional Hazards and Random Survival Forests are adopted for modelling the time-to-failure and the associated survival functions. Historical telematics data from electric heavy-duty vehicles in operations are used to train the models. Feature selection and hyperparameter tuning techniques are implemented to improve the model performance.

Keywords: Control of systems in vehicles, Guidance, navigation and control of vehicles
Abstract: The paper addresses a problem of constrained spacecraft attitude stabilization with simultaneous reaction wheel (RW) desaturation. The spacecraft has a reaction wheel array (RWA) consisting of four RWs in a pyramidal configuration. The developments exploit a spacecraft dynamics model with gravity gradient torques. The linearized dynamics are shown to be controllable at almost all RWA configurations. Configurations that result in the highest Degree of Controllability are elucidated. A strategy that combines an incremental reference governor and time-distributed model predictive control is proposed to perform constrained RW desaturation at low computational cost. Simulation results of successful RW desaturation maneuvers subject to spacecraft pointing constraints, RW zero-speed crossing avoidance and limits on control moments are reported.

Keywords: Space exploration and transportation, High accuracy pointing, Extremum seeking and model free adaptive control
Abstract: This paper addresses the pointing acquisition phase of the Laser Interferometer Space Antenna (LISA) mission as a guidance problem. It is formulated in a cooperative game setup, which solution is a sequence of corrections that can be used as a tracking reference to align all the spacecraft' laser beams simultaneously within the tolerances required for gravitational wave detection. We propose a model-free learning algorithm based on residual-feedback and momentum, for accelerated convergence to stable solutions, i.e. Nash Equilibria. Each spacecraft has 4 degrees of freedom, and the only measured output considered are laser misalignments with the local interferometer sensors. Simulation results demonstrate that the proposed strategy manages to achieve absolute misalignment errors <1 microrad in a timely manner.

Keywords: Space exploration and transportation, Guidance, navigation and control of vehicles, Autonomous systems
Abstract: Future interplanetary missions will carry more and more sensitive equipment critical for setting up bases for crewed missions. The ability to manoeuvre around hazardous terrain thus becomes a critical mission aspect. However, large diverts and manoeuvres consume a significant amount of fuel, leading to less fuel remaining for emergencies or return missions. Thus, requiring more fuel to be carried onboard. This work presents fuel-optimal guidance to avoid hazardous terrain and safely land at the desired location. We approximate the hazardous terrain as step-shaped polygons and define barriers around the terrain. Using an augmented cost functional, fuel-optimal guidance command, which avoids the terrain, is derived. The results are validated using computer simulations and tested against many initial conditions to prove their effectiveness.

Keywords: Decision making and autonomy, sensor data fusion, Guidance, navigation and control of vehicles, High accuracy pointing
Abstract: To push the boundaries of autonomy in space, the spacecraft must rely on its own sensors to achieve positioning and environmental perception. In this context, the key problem of autonomous navigation is the nonlinear state estimation of the spacecraft in a dynamic 3D environment. In this paper, we propose a new approach based on a single-state sub-partitioning of the state vector and a partial updating of the vector of weights according to the specific information provided by each sensor. In this way, we avoid to lose information in the resampling phase thanks to a parallelization approach. The proposed method has been applied to an Earth observation mission and the efficacy of the proposed approach is demonstrated with a numerical example using a high-fidelity orbital simulator.

Keywords: Space exploration and transportation, Guidance, navigation and control of vehicles, Neural networks
Abstract: An integrated guidance and control system for tracking a reference flight path trajectory of a reusable launch vehicle in the presence of model uncertainties, external disturbances, and measurement noises is proposed. To achieve this goal, a properly-trained feedforward neural network to estimate the uncertainties in conjunction with a disturbance observer, which estimates external disturbances and the estimation error made by the neural network, is implemented in the scheme. In addition, a state observer is designed whose state estimation error is included (along with the tracking error) in the learning algorithm (of both the neural network and the disturbance observer); such learning is called composite learning. The combined neural network and state observer (employed in both the guidance and control loops) can consistently compensate for different complex and/or uncertain terms in the dynamic model of a reusable launch vehicle. More specifically, the introduced design results in an asymptotic tracking of the reference command, thereby providing a resilient flight control system. Extensive simulations are performed to show the effectiveness and robustness of the proposed integrated guidance and control system for trajectory tracking in the presence of uncertainties, aerodynamic disturbances, and measurement noises.

Keywords: Autonomous systems, Multi-agent systems
Abstract: A distributed formation control scheme is proposed for achieving a desired formation of multiple spacecraft in orbit around the Earth within a free will arbitrary settling time, over an undirected communication network. Using the Clohessy-Wiltshire equations as relative position dynamics model for the spacecraft in a circular reference orbit around the Earth, the proposed control strategy is able to achieve the desired formation of the spacecraft using only relative position measurements. Simulations are presented to demonstrate the effectiveness of the proposed strategy.

Keywords: Mechatronics, Identification and control methods, Mechatronic systems
Abstract: This paper describes the development and implementation of three driving strategies for the oscillation amplitude control of a resonant rotational reluctance actuated scanning mirror system. The three driving strategies exploit control of current input amplitude, pulse duration, and the phase between the input and the mover angle. Linear PI-controllers are developed for each driving strategy for the linearized plant around an operating point of the highly nonlinear system. The implemented driving strategies are compared over a wide operating range of the scanning mirror system, regarding their closed-loop dynamics in the time-domain, revealing their fundamental dierences. Phase-control is the most promising control strategy, delivering the highest closed-loop bandwidth of at least 11 Hz over the entire investigated operating range at worst with a rise-time of 30 ms at a step response.

Keywords: Identification and control methods, Micro and nano mechatronic Systems, Nonlinear system identification
Abstract: Magnetic microrobot (MM) is promising for in-body therapy due to its advantage of untethered controllability. Since the motion of MMs can be directly manipulated by the magnetic field (MF) created by electromagnetic coil systems, it is essential to study the MF generation control methods. In this study, iron core-based solenoids are employed in an electromagnetic coil system that can produce a stronger MF than regular coil systems. However, the coil system presents hysteretic nonlinearity due to the iron core's coercivity, which degrades closed-loop control performances. To tackle this challenge, an effective hysteresis compensation method is developed to linearize the solenoid system. Specifically, the least squares support vector machine (LSSVM) is adopted to model the hysteretic behavior of the solenoid system, and the discrete empirical interpolation method (DEIM) is applied to reduce complexity of the model for enhancing calculation efficiency. Based on the LSSVM and DEIM approaches, a highly efficient LSSVM-DEIM hysteresis compensator is developed. In order to generate precise dynamic MF effectively, the LSSVM-DEIM hysteresis compensator is combined with prevalent feedback controllers to form composite control schemes. Experiments show that the control system with the LSSVM-DEIM hysteresis compensator can produce a more accurate MF compared with that of the benchmark control methods.

Keywords: Identification and control methods, Application of nonlinear analysis and design, Motion control systems
Abstract: The paper is devoted to the control of electromagnetic actuators. Compared to most results in the literature, the design explicitly takes into account the nonlinearities and the magnetic saturation. An approach based on singularly perturbed systems is proposed to take advantage of the different time scales of the model and provide a minimal complexity control law. Associated with an integral action, this control law ensures accuracy and robustness w.r.t. some disturbances and some unknown parameters of the actuator. The use of the singular perturbation method allows also to limit the implementation complexity compared to classical methods such as backstepping control. The synthesis of the control law is then tested through simulations that illustrate the efficiency of the method.

Keywords: Mechatronics, Identification and control methods, Motion control systems
Abstract: The increasing demands for motion control result in a situation where Linear Parameter-Varying (LPV) dynamics have to be taken into account. Inverse-model feedforward control for LPV motion systems is challenging, since the inverse of an LPV system is often dynamically dependent on the scheduling sequence. The aim of this paper is to develop an identification approach that directly identifies dynamically scheduled feedforward controllers for LPV motion systems from data. In this paper, the feedforward controller is parameterized in basis functions, similar to, e.g., mass-acceleration feedforward, and is identified by a kernel-based approach such that the parameter dependency for LPV motion systems is addressed. The resulting feedforward includes dynamic dependence and is learned accurately. The developed framework is validated on an example.

Keywords: Identification and control methods, Mechatronics for mobility systems, Modeling
Abstract: For most electrified railways, pantographs play a vital role in transmitting energy from overhead line to vehicles, and therefore a stable and continuous contact behavior is required. This paper proposes a disturbance observer-based sliding mode controller (DO-SMC) for the problem of pantograph–catenary contact force regulation. The simulation results show that the DO-SMC with a chattering alleviation approach can effectively reduce contact force fluctuation through a reasonable control input.

Keywords: Identification and control methods, Motion control systems, Mechatronics
Abstract: This paper presents a position estimation compensation method considering dq-axis cross-coupling factors of Interior Permanent Magnet Synchronous Motors (IPMSMs) inductance for position-and-torque-sensorless admittance control. The cross-coupling factors are from dq-axis mutual inductance and deteriorate control and position estimation performance in control methods without the model. Since dq-axis cross-coupling factors of inductance vary with the rotor position and the current, measurement and estimation of the values are difficult. The proposed method estimates the cross-coupling factors and position estimation errors using dq-axis current variation from the injected voltage. In addition, the proposed method compensates the estimated position based on a voltage equation, including the dq-axis cross-coupling factors, by adding the estimated error. As a result, the vibration of current, velocity, and estimated reaction torque were reduced, and the performance of a position-and-torque-sensorless admittance control system was improved. The experimental results show the proposed method's validity and response by the sensorless admittance control.

Keywords: Discrete event modeling and simulation, Supervisory control and automata, Reachability analysis, verification and abstraction of hybrid systems
Abstract: In this discussion paper, we highlight a synthesis-based engineering (SBE) framework for the engineering of supervisory controllers. The engineering of supervisory controllers for large and complex cyber-physical systems requires dedicated engineering support. The Compositional Interchange Format (CIF) language and toolset have been developed for this purpose. CIF is part of the Eclipse Supervisory Control Engineering Toolkit (ESCET) project. In this paper we discuss language and tool aspects that are necessary for industrial application, but under-exposed in academic research efforts.