Research On Multi-objective Control Strategy Of Thermal Management System For Pure Electric Vehicle Under Low Temperature Conditions

At present,pure electric vehicles have become an important direction for the development of the automotive industry.However,under the low-temperature extreme conditions,the battery performance decays seriously due to the decreasing activity and increasing internal resistance of the pure electric vehicle power battery,and the heating energy consumption of the passenger compartment also increases significantly,which leads to a significant decrease in the economy and power of the whole vehicle and makes it difficult to open the market of pure electric vehicles in the high-altitude and high-latitude areas,so it has become an urgent need to study the key technologies of pure electric vehicle thermal management.As one of the core technologies of thermal management system,the control strategy can directly affect the performance of thermal management system and even the whole vehicle.Therefore,this paper focuses on the key control technology of integrated thermal management system of pure electric vehicles under low temperature extreme conditions,aiming to improve the economy,power and passenger compartment comfort of the whole vehicle.Since the integrated thermal management system has complex and non-linear characteristics,this thesis proposes an intelligent control policy based on Q-Learning algorithm to enhance the adaptability of thermal management system control strategy to different operating conditions.The main research conducted based on this is as follows:（1）Establish an integrated thermal management system and vehicle dynamics model.Firstly,establish mathematical models for power battery thermal management,passenger compartment thermal management,and vehicle dynamics.Then,a vehicle simulation model is built based on the Amesim and MATLAB/Simulink joint simulation platform.Finally,a real vehicle test was conducted in an environmental warehouse,and the test data was compared with the simulation data.The comparison results showed that the error between the simulation data and the test data was within 6.21%,and the simulation model met the design requirements.（2）Develop a vehicle thermal management control strategy considering different battery heating cut-off temperatures.Firstly,introduce the control process of the battery circuit and passenger compartment circuit,and develop control strategies for the thermal management system based on different battery heating cut-off temperatures.Then,a multi-objective evaluation function is established to evaluate the control effectiveness of different strategies,considering the overall vehicle economy,power performance,and passenger compartment comfort.Finally,through simulation and verification of WLTC（World Light Vehicle Test Cycle）and NEDC（New European Driving Cycle）operating conditions,two conclusions were obtained:heating the power battery in an environment of-20℃resulted in an economic improvement of over 8%in both operating conditions,with a significant increase in power performance.This confirms the necessity of heating the power battery;The optimal heating cutoff temperature for power batteries varies under different operating conditions.The optimal heating cutoff temperature for batteries under WLTC operating conditions is 10℃,and the optimal heating cutoff temperature for batteries under NEDC operating conditions is 5℃.（3）According to the known driving conditions,an integrated thermal management system control strategy based on dynamic programming（DP）optimization algorithm is proposed.Firstly,the basic principle of DP algorithm was explained and its calculation process was analyzed.Then,with the comprehensive optimization of vehicle economy,power performance,and passenger compartment comfort as the goal,an optimization model is established with battery SOC（State of Charge）,battery temperature T_Bat,and passenger compartment temperature T_Cbn as state variables,and battery heating power P_Bat and passenger compartment heating power P_Cbn as control variables.A control strategy for the vehicle thermal management system based on DP algorithm is developed.Finally,simulation verification was conducted on the WLTC and NEDC operating conditions,and the results showed that the DP strategy improved the overall performance by 9.92%and 11.93%compared to the rule strategy under both operating conditions,respectively;There are significant differences in the control strategies of thermal management systems under different operating conditions,and integrated thermal management control strategies can be considered for different operating conditions.（4）A control strategy for an integrated thermal management system based on Q-learning optimization algorithm is proposed for unknown driving conditions.Firstly,the basic principle of Q-learning algorithm was introduced,and the calculation process of strategy iteration algorithm was analyzed.Then,establish the probability matrix of demand power state transition for WLTC and NEDC operating conditions.Secondly,with the comprehensive optimization of vehicle economy,power performance,and passenger compartment comfort as the goal,an optimization model is established with demand power P_Req,battery SOC,battery temperature T_Bat,and passenger compartment temperature T_Cbn as state variables,and battery heating power P_Bat and passenger compartment heating power P_Cbn as control variables.A control strategy for the vehicle thermal management system based on Q-learning algorithm is developed.Finally,the simulation results of the Q strategy were compared with the rule strategy and DP strategy,and the comparison results showed that the Q learning strategy has strong adaptability to different working conditions.