Predictive energy management for plug-in hybrid electric vehicles considering electric motor thermal dynamics

Abstract

Energy management is essential for improving the fuel economy of plug-in hybrid electric vehicles (PHEVs). Some existing efforts have focused on optimizing fuel consumption and battery degradation, but without adequately considering the onboard electric motor's (EM) thermal dynamics. To address this research gap, this paper proposes a predictive energy management strategy considering EM thermal control. Specifically, we make three main contributions that distinguish our study from the existing studies. First, we design four velocity predictors based on the artificial neural network (ANN) and examine their prediction accuracy and computational efficiency. Second, we present a Pontryagin's Minimum Principle-based model predictive control (PMP-MPC) framework that includes EM thermal dynamics. The framework minimizes the operating costs while ensuring that the EM temperature is less than the limit value. Finally, we analyze and compare the effects of different reference temperature thresholds and preview horizon sizes on the fuel economy and EM temperature. The results demonstrate that the proposed PMP-MPC approach can effectively control the EM temperature rise and realize online applications with high computational efficiency.