Real-time model predictive control for thermal management in an automotive ICE with an electrified cooling system

The electrified cooling system decouples the actuators from the internal combustion engine (ICE), permitting advanced control for fuel saving. Almost all current thermal management control strategies only focus on the cooling system, which improves the fuel economy by thermal chain rather than by electric and dynamic chain. This paper elevates the control strategy from the cooling system level to the power system level. First, based on the mechanismelectricity-heat integrated optimisation, a minimum fuel consumption control strategy is proposed. Then an efficient solving method that combines Pontryagin's minimum principle (PMP) and bisection method is proposed, realising the online model predictive control (MPC). Simulation results show an improved fuel saving effect of the proposed control strategy compared to traditional cooling system over various driving cycles, and the high computational efficiency of the proposed solving method compared to the traditional optimisation methods. The MPC implementation on a hardware platform confirms the real-time feasibility.