Research On Energy Management Strategy Of PHEV Based On Adaptive Dynamic Programming

Plug-in hybrid electric vehicle(PHEV)can not only use electric mode to achieve green travel on urban roads and reduce pollutant emissions,but also has the characteristics of traditional fuel vehicles.When the power is insufficient,it can be driven by an engine,which effectively avoids the problems of insufficient energy and short cruising range.How to coordinate the distribution of motor driving torque and engine torque and effectively reduce the fuel and exhaust emissions of the whole vehicle has become an important content of PHEV energy management strategy research.In order to better coordinate the energy distribution between the motor and the engine,and to reduce the fuel consumption and exhaust emissions of the whole vehicle and stabilize the SOC of the battery pack,a PHEV energy management control strategy based on ADP is studied and designed.The main research contents are as follows:(1)Mathematical analysis and model establishment of vehicle power system.The torque and power distribution modes of PHEV in various operation modes and the energy flow in various operation modes are analyzed.The dynamic performance and mathematical model of the whole vehicle are analyzed in detail,and the dynamic system model of the whole vehicle is established according to the dynamic performance index and basic parameters of the whole vehicle.(2)Aiming at the problem of PHEV energy management,the control strategy of ADHDP is studied and designed.An energy management controller based on action-dependent heuristic dynamic programming(ADHDP)is designed by using a three-layer BP neural network.Through the learning and training of ADHDP neural network,the optimal control quantity of engine output torque can be finally obtained.The feasibility is verified by the joint simulation of MATLAB/Simulink and ADVISOR 2002,and compared with the fuzzy control strategy.The results show that the battery SOC value of ADHDP control strategy can be stabilized above 0.60 after one,two and three cycles,and the fuel saving effect of the whole vehicle is remarkable,and the exhaust emissions are obviously reduced.(3)In order to further reduce fuel consumption and exhaust emissions,DHP control strategy and Gr HDP control strategy are studied and designed respectively.Aiming at the problems of nonlinearity,complexity,strong interference and uncertainty in the vehicle drive system,three-layer BP neural network is used to design the control strategies of double heuristic dynamic programming DHP and goal representation heuristic dynamic programming Gr HDP respectively.The comparative simulation results show that among the three ADP control strategies,Gr HDP energy management control strategy has a better control effect.While ensuring the relative balance of the SOC of the vehicle battery pack,this control strategy enhances the endurance,and has a very obvious effect on improving the fuel economy and environmental protection of the vehicle.