A driving cycle recognition-based optimized fuzzy energy management strategy for enhanced performance of parallel hybrid electric vehicles

Abstract

The energy management strategy (EMS) is pivotal in hybrid electric vehicles (HEVs), critically influencing fuel economy and dynamic performance. Conventional EMS approaches often lack adaptability under complex and variable driving conditions. To address this limitation, this study proposes a novel dual-layer adaptive EMS framework that integrates driving cycle recognition (DCR) and an optimized fuzzy logic controller (FLC). The upper layer employs a back propagation neural network (BPNN) for real-time identification of driving conditions. Based on the recognized driving cycle category, the lower layer dynamically selects and applies an corresponding optimized fuzzy control strategy to achieve adaptive power distribution, thereby improving fuel efficiency and overall system performance across diverse environments. The simulations demonstrate that the DCR model attains a high recognition accuracy of 98.6%. Compared to rule-based EMS, conventional FLC, and particle swarm optimization (PSO)-optimized FLC strategies, the proposed DCR-PSO-FLC approach reduces fuel consumption by 13.18%, 5.56%, and 3.62%, respectively. Additionally, it achieves a 5.07% improvement in average thermal efficiency over the baseline fuzzy rule-based strategy. These results validate the superior energy efficiency and enhanced adaptive control capability of the proposed EMS.