Holistic energy management strategy for hybrid electric heavy-duty vehicles based on proximal policy optimization with the consideration of cabin temperature comfort

Electric vehicles (EVs) are becoming acknowledged as sustainable and environmentally beneficial alternatives. Hybrid electric vehicles (HEVs) have emerged as a feasible way to mitigate “range anxiety.” Many energy management strategies (EMS) for EVs neglect the energy consumption of air conditioning systems (ACS), leading to suboptimal energy utilization. This study tackles this gap by concentrating on hybrid electric buses (HEBs) and incorporating ACS requirements into their energy management system for enhanced powertrain efficiency. A control-centric cabin thermal management model is integrated into the powertrain structure. A customized drive cycle characterized by mixed city-highway circumstances and minimal traffic density is generated using Simulation of Urban Mobility (SUMO) software to produce diverse training data. The Proximal Policy Optimization (PPO) algorithm is utilized to improve EMS performance, contrasting PPO-comfort EMS against the Twin Delayed Deep Deterministic Policy Gradient (TD3) approach and TD3-Bang-Comfort strategy. The study incorporates a cost analysis to evaluate the economic advantages of the suggested EMS. The findings demonstrate that the proposed EMS optimizes training stability and convergence by 11.90 % and 18.65 % than other strategies, substantially lowers operational expenses, and improves cabin thermal comfort, resulting in cost reductions of 10.76 % and 16.03 % in comparison to the TD3-comfort and TD3-Bang-comfort strategies.