Model Predictive Control-based Engine Idle Speed Regulation with Various Coordinated Controls Using an Instantaneous Engine Model

This article presents the development of coordinated control of throttle, spark advance, and variable valve timing (VVT) in a model predictive control (MPC) framework for engine idle speed control application. The objective in this work is to develop an idle speed controller, which can maintain desired engine idle speed with a fast response while being subjected to load torque disturbances (e.g., HVAC, auxiliary loads), system nonlinearities (e.g., intake to torque production delay), and the like. The proposed controller’s efficacy is demonstrated on practically validated instantaneous crank angle-based engine model. For the development of the controller, the equivalent linearized plant models are derived using system identification from the input and output data of the practically validated instantaneous crank angle-based engine model. Present states’ information required by the controller is estimated using a Kalman filter. To assess the performance of the developed multivariable MPC controller, a weighted PID controller is also developed as a benchmark and their comparative simulation results are analyzed and presented along with corresponding experimental results from a real engine.