Predictive energy management strategy for connected 48V hybrid electric vehicles

Abstract

The challenges encountered in the development of a predictive energy management strategy (EMS) for hybrid electric vehicles (HEVs) include improving the vehicle speed prediction accuracy and resolving the contradiction between the real-time performance and optimality. Connected vehicles provide possible vehicle speed prediction accuracy improvements, further optimizing the EMS. Therefore, this paper proposes a novel predictive EMS for connected vehicles. Based on traffic information, the vehicle speed trajectory is predicted according to driver's intention inference and prediction models for each intention. Since the A* algorithm can quickly find the shortest path in the road network under the guidance of the distance from the goal, this algorithm is introduced to search for the optimal EMS within the prediction horizon under the guidance of a proposed heuristic function. Then, the searched EMS is fused with an offline-optimized strategy to improve the real-time performance. Taking a 48 V HEV as an example, the predictive EMS is simulated using traffic data generated by traffic simulation software, and the fuel economy is only 1.16% lower than that under the global optimal strategy. Finally, hardware-in-the-loop tests are performed to verify the real-time performance.