A cloud energy management strategy for intelligent connected HEVs based on end-edge-cloud collaboration

Intelligent connected hybrid electric vehicle (ICHEV) is one of the important means to achieve the energy saving and carbon neutralization in the future. For the large-scale application of ICHEVs, a vehicle-road-cloud collaborative cloud energy management method is proposed. First, an end-edge-cloud three-layer architecture for ICHEVs energy management system is presented, and a cloud energy management strategy (CEMS) based on the cooperation of vehicle-road end, edge cloud and central cloud is proposed. Then, an improved deep learning model of stacked denoising autoencoder (I-SDA) is constructed to predict the traffic speed of a single road section, and a full-trip vehicle speed prediction algorithm called multi-step spatio-temporal collaboration (MSSTC) algorithm is proposed for the central cloud. Next, the energy management is modeled as a multi-stage optimized decision problem, and a general DP solution algorithm is designed to obtain the state of charge (SOC) and torques of the vehicle. On this basis, the optimal full-trip SOC trajectory acquisition algorithm (OSocTA) for the central cloud and the optimal torque distribution algorithm (OTD) for the edge cloud are proposed. The final experiments show that compared with the traditional charge depleting charge sustaining (CDCS) strategy, the fuel consumption of the CEMS proposed in this paper is reduced by about 48%, which can better meet the needs of building cost-effective vehicle energy management in the future large-scale applications of ICHEV.