An efficient multi-objective hierarchical energy management strategy for plug-in hybrid electric vehicle in connected scenario

Abstract

Nowadays, the comprehensive performance of plug-in hybrid electric vehicle (PHEV) is expected to be further improved with development of connected vehicle technology. However, the strong coupling and traffic flow uncertainty characteristics of connected scenario pose formidable challenge to existing energy management strategies (EMSs) in terms of optimization effect and computational efficiency. For comprehensively improving connected PHEV performances including energy saving, safety, traffic efficiency and computational efficiency, a multi-objective hierarchical EMS with less computational burden is proposed by incorporating resistance network (RN) triggered motion planning and alternating direction method of multipliers (ADMM) based convex torque optimization. Specifically, the RN method is employed to characterize and decouple the complex interaction relationship within connected scenario from internal mechanism perspective, enabling the velocity profile optimization issue that with fixed end time constraint and online correction mechanism for traffic flow uncertainty. According to the velocity profile optimized in cloud level, the convex formulation of model predictive control (MPC) based torque distribution problem is formulated in vehicle level, and an efficient ADMM algorithm is used for its solution, with the aim of satisfying energy saving and practical application requirements simultaneously. Based on real connected information, the superiorities of proposed EMS are verified by both simulation and hardware-in-loop (HIL) experiment.