Research On Fuzzy Optimal Energy Management Of Electric Vehicles By Combining The Construction And Prediction Of Driving Cycle

At present,the range of battery electric vehicles(BEVs)is severely limited by the development of battery technology.Therefore,how to improve the energy utilization of BEV has become a key issue under the premise of limited battery energy.Meanwhile,the driving cycle has huge effect on the effectiveness of the energy management strategy.Most of the current energy management strategies take the standard driving cycles as input that ignore the driving characteristics of different cities.Hence,in this paper,a BEV with single power supply is selected as the research object.And then,considering the influence of urban driving cycle and air-conditioning load on energy management,study on fuzzy optimal energy management of electric vehicles by combining the construction and prediction of driving cycle has been carried out as follows:(1)The urban driving cycle is constructed,taking a BEV as test vehicle.A large number of urban driving data have been collected,and the kinematic segments are divided into four clusters by the principal component analysis and K-means clustering algorithm.Then,the urban driving cycle is constructed based on the representative segments,which are selected according to the time proportion of each cluster and the distance between kinematic segments and the cluster center.Finally,the effectiveness of the constructed driving cycle has been verified by comparison with the preprocessed data.(2)A combined driving cycle prediction method is proposed.In order to verify the Markov property of driving cycle,the speed correlation analysis is carried out on the current and adjacent driving cycle data.The size of rolling time window is determined with the prediction accuracy as the evaluation index.The rolling prediction method is used to predict the vehicle speed,and the state transition matrix based on the constructed driving cycle is applied to predict the vehicle speed in the first rolling time window.This method can not only guarantee that the speed can be predicted at the initial time of vehicle operation,but also guarantee the real-time and accuracy of the prediction.(3)The vehicle model of BEV is established.The driver model is built based on the PID controller to reflect the intention of the driver.The battery model and motor model are built using the test data.And the vehicle longitudinal dynamics model is built to obtain the driving resistance and the actual vehicle speed.The air-conditioning model is built to calculate the energy consumption of the air-conditioning.Moreover,vehicle control unit model is built to calculate the vehicle energy consumption and determine the energy allocation.These work lays a foundation for the subsequent simulation analysis of the energy management strategies.(4)The simulation analysis of energy management strategy based on driving cycle prediction is completed.The simulation results of logic threshold control strategy,logic threshold control strategy based on driving cycle prediction and fuzzy logic control strategy based on driving cycle prediction are compared in terms of SOC,current,voltage and energy consumption rate.And advantages of the fuzzy logic control strategy based on the driving cycle prediction are verified.Then,the membership functions of the fuzzy controller are optimized by applying the grey wolf optimization algorithm,and the simulation results show that the optimal strategy can effectively reduce the energy consumption rate.(5)The hardware in-loop test of energy management control strategy is carried out.The vehicle controller algorithm software by means of rapid prototype development platform is developed,and the underlying module,hardware I/O interface and communication module are selected and configured.Then,the algorithm is compiled and brushed to the controller.The hardware in-loop test platform is built,and the test results indicate the effectiveness and feasibility of the proposed fuzzy optimal energy management strategy in real time environment.