Experimental comparisons between LQR and MPC for spark-ignition engine control problem

Abstract

This paper proposes an experimental comparison between two optimal controllers for automotive spark ignition engines, including an gain-scheduled linear quadratic regulation (LQR) controller and model predictive controller (MPC). The aim of the study is to highlight the control effects between the LQR and MPC scheme on the specific engine control problem, and provide a reference for the future controller design. The control problem is formulated to ensure the fast torque tracking performance, and meanwhile to improve the thermal efficiency by reducing the pumping loss. The nitrogen oxide emission is chosen as constraint. The experiment were implemented on the full-scale gasoline engine, and the experiment results demonstrate the performance of the proposed optimal control designs.