Ahmed, Omar (University of Michigan), Middleton, Robert (University of Michigan), Tran, Vivian (University of Michigan), Weng, Andrew (University of Michigan), Stefanopoulou, Anna G. (Univ Of Michigan), Kim, Kenneth (DEVCOM Army Research Laboratory), Kweon, Chol-Bum (DEVCOM Army Research Laboratory)

Model Predictive Control of Combustion Phasing in Compression Ignition Engines by Coordinating Fuel Injection Timing and Ignition Assist

Scheduled for presentation during the Regular Session "Modeling, Estimation, and Control of Internal Combustion Engine- II" (MoBT3), Monday, August 29, 2022, 15:50−16:10, Pfahl Hall 140

10th IFAC International Symposium on Advances in Automotive Control, August 28-31, 2022, Columbus, Ohio, USA

This information is tentative and subject to change. Compiled on April 24, 2024

Abstract

Internal combustion engines may use ignition assisting heating elements such as glow plugs to facilitate combustion control in automotive or aircraft powertrains that operate with synthetic fuels of varying ignition behavior or at extreme inlet conditions. This work presents a model predictive controller (MPC) that regulates combustion phasing in compression ignition engines on a cycle-to-cycle basis by coordinating fuel start of injection (SOI) with power supplied to a glow plug acting as an ignition assist (IA) device, while enforcing IA actuator range and rate constraints. Simulations were conducted using a nonlinear virtual engine informed by data from a commercial engine operating at a condition that induced high combustion variability. A rate-based MPC formulation leveraging state estimate feedback and integral setpoint tracking was developed. Simulation results show the MPC scheme ensures steady-state tracking of combustion phasing within 70 engine cycles, conserves IA usage whenever possible to reduce thermo-mechanical stress on the actuator, and maintains closed-loop combustion variability at only 4% higher than the open-loop system variability. Furthermore, the controller maintains reference tracking even if combustion sensitivity to the actuators deviates by more than 20% from the controller's internal model, without the need for retuning control parameters.