Energy management strategy for electro-hydraulic hybrid electric vehicles considering optimal mode switching: A soft actor-critic approach trained on a multi-modal driving cycle

Hybrid electric vehicles (HEVs) feature multiple working modes. Thoughtful selection of these modes can optimally balance driving performance, power demands, and energy consumption, thereby enhancing the overall efficiency of the vehicle. This paper presents a soft actor-critic (SAC) approach trained on a multi-modal driving cycle (MDC) for selecting operational modes of electro-hydraulic hybrid electric vehicle (EHHEV). Firstly, characteristic parameters are extracted and clustered for five typical driving cycles through principal component analysis and K-means clustering, creating a multi-modal driving cycle. Secondly, based on the operational characteristics of EHHEV, state variables, action variables, reward functions, learning rates, and other parameters are set for the SAC algorithm, and the EMS framework is built based on the electro-hydraulic hybrid electric power system. Subsequently, the SAC algorithm is trained using the MDC to construct the SAC-MDC EMS. Results demonstrate that compared to EV, RB EMS, and SAC EMS, IREC achieves maximum improvements of 22.38 %, 5.55 % and 0.80 %, respectively. The dynamic performance and the motor load optimization capability are also enhanced. To further validate the practicality and reliability of the SAC-MDC EMS, this paper validates it using actual driving data, revealing that it still exhibits outstanding performance.