Research On Co-optimization Strategy Of Velocity Planning And Energy Management For Connected Hybrid Vehicle Fleet

Hybrid Electric Vehicle(HEV)has enormous potential in energy conservation and emission reduction.The development of intelligent connected technology in vehicles has brought the possibility of further improving traffic efficiency and fuel economy.In addition,vehicle queue driving has also shown unique advantages in energy conservation and safety improvement.The connected HEV fleet can demonstrate better performance in terms of power,fuel economy,safety,and comfort.However,the research and design of vehicle control strategies for connected HEV fleet also pose significant challenges,especially the mutual influence of HEV velocity planning and energy management.Therefore,this dissertation focuses on the co-optimization research of velocity planning and energy management for connected HEV fleet.Based on the establishment of HEV model and road scenario model,co-optimization strategies for velocity planning and energy management based on hierarchical optimization and synchronous optimization are designed for the pilot vehicle and the following vehicles,respectively.The main research content is as follows:Firstly,key components of HEV and road scene models were established.Analyzed the structure of the HEV power system,combined with the working characteristics of the power coupling mechanism,studied the speed and torque coupling relationship between the engine and the dual motor under different working modes,and constructed key component models such as the engine,motor,power coupling mechanism,power battery,and main reducer,laying the foundation for the research of energy management strategies;Set the road scene and signal phase and timing(SPa T)information,and construct a road scene model,laying the foundation for the design of velocity planning optimization strategies.Secondly,a co-optimization strategy for velocity planning and energy management of pilot vehicle based on hierarchical optimization was proposed.Considering the complexity of the co-optimization problem of velocity planning and energy management for the pilot vehicle in macro scenarios,a hierarchical control-based co-optimization architecture is proposed.The upper layer constructs a multi-objective optimization problem considering vehicle fuel economy,traffic efficiency,and safety based on signal phase and timing information,and obtains the optimal vehicle speed trajectory based on simulated annealing algorithm(SA);At the lower level,an energy management strategy based on model predictive control(MPC)is designed based on the optimal vehicle velocity trajectory to distribute power between the engine and motor.The upper and lower levels collaborate to improve the traffic efficiency and fuel economy of HEV.Thirdly,a co-optimization strategy for velocity planning and energy management of following vehicle based on synchronous optimization was proposed.Combining the advantages of the model free deep reinforcement learning(DRL)algorithm,a co-optimization strategy of the following vehicle based on DRL for velocity planning and energy management is proposed.The corresponding state space and action space are designed,and a multi-objective reward function is designed based on the requirements of the following vehicle for fuel economy,safety,and comfort.The proposed co-optimization strategy based on Soft Actor Critical(SAC)algorithm is designed with network,hyperparameter and simulation test parameters.The training results of the algorithm are further analyzed,and the convergence of the training is proved.Finally,a simulation validation study was conducted on the co-optimization strategy of velocity planning and energy management for connected HEV fleet.The velocity planning simulation results and energy management strategy simulation results of the pilot vehicle were analyzed,and the fuel economy improvement reached 39.3%;The co-optimization simulation results of the following vehicle were analyzed from aspects such as smoothness,safety,and fuel economy.Compared with the comparison strategy,the fuel economy of the following vehicle increased by 6.0% and 8.0%,respectively;Finally,a comparative analysis was conducted on the overall results of the co-optimization simulation of velocity planning and energy management for the connected hybrid electric vehicle fleet.The average equivalent fuel consumption of the fleet was reduced by 41.2%,verifying the superiority of the proposed strategy in this dissertation.