Study On Adaptive Energy Management Strategy For Extended-range Electric Vehicle Based On Micro-trips Identification

It is one of the effective solutions to solve current exhaust emission pollution,oil crisis and energy security by developing extended-range electric vehicle(EREV).The energy management strategy can improve fuel economy while meeting the power demand by reasonably distributing the driving power among multiple power sources.In this paper,the adaptive energy management strategy based on micro-trips identification is studied for EREV.The main contents include:The forward simulation model of EREV is established to lay the foundation for the follow-up research.After comparing the simulation results of thermostat energy management strategy,DP energy management strategy is selected as the reference for subsequent optimization.The simulation results of A-ECMS based on PI and BP neural network are compared.A-ECMS strategy based on PI is using PI control to adjust the equivalent factor of Equivalent Consumption Minimization Strategy(ECMS)to improve the robustness of ECMS.A-ECMS strategy based on BP neural network takes the optimization result of DP strategy as the following target to obtain the instantaneous optimal equivalent factor of ECMS.Then the demanded power of vehicle and SOC of battery are used as input terms,and the instantaneous optimal equivalent factor of corresponding ECMS strategy is used as output term to train BP neural network.The K-means++clustering algorithm is used to cluster the micro-trips and construct four typical driving cycles.Three characteristic parameters are selected by using the correlation between characteristic parameters and fuel consumption.Comparing the accuracy of the decision tree algorithm and the distance discriminant method,the distance discriminant method with higher accuracy is selected as the micro-trips identification algorithm.The micro-trips identification algorithm and A-ECMS strategy are combined for simulation verification.The results show that algorithm of A-ECMS using BP neural network to optimize the equivalent factor is close to the result of DP strategy through micro-trips identification,which has the best effect on improving the fuel economy of EREV under unknown driving cycles.