Research On Electric Vehicle Range Estimation Method Based On Battery Status And Driving Energy Consumption Prediction

In the context of the global energy crisis and the transformation of green low-carbon economy,the transformation of the automobile industry to electrification has become a general trend.Compared with traditional fuel vehicles,electric vehicles have the advantages of no pollution and low noise.However,there are also problems such as limited driving range,long charging time and inconvenient charging,which easily lead to the phenomenon of “mileage anxiety” of electric vehicle drivers.This phenomenon directly reducing the driver’s driving experience and restricting the further development of the electric vehicle industry.Therefore,accurately predicting the driving range of electric vehicles and providing drivers with real-time information on the remaining driving range are of great significance for guiding drivers to rationally plan travel routes,alleviating drivers’ mileage anxiety and promoting the development of the electric vehicle industry.Firstly,the original data of the collected electric vehicle operation is detected,the data fields related to the research in this thesis are extracted,the data of the charging stage and the driving stage of the electric vehicle are screened and separated,the problems of the data are analyzed,the outliers are filtered and the missing values are filled,and the data slicing rules are established to segment the charging data and the driving data,so as to provide high-quality experimental data for subsequent modeling.Secondly,the characteristic parameters of the electric vehicle battery state of charge and the battery state of health estimation model are extracted.The characteristic variables related to the target variables are selected.A battery state estimation model based on the XGBoost algorithm is established.The particle swarm optimization algorithm is used to optimize the parameters of the model.The training set and the test set data are used to train and test the model respectively.The estimation effect of the model is demonstrated,the error coefficient of the model is calculated,and the prediction effect is compared with the battery state estimation model based on Light GBM algorithm to verify the accuracy of the electric vehicle battery state estimation model based on PSO-XGBoost algorithm.Thirdly,the characteristic variables of the driving condition prediction model are extracted,and the driving condition prediction model based on GRU neural network is established.The data of different driving conditions are selected to test the prediction accuracy of the model,The prediction effect of the model is demonstrated,the error coefficient of the model is calculated,and the prediction effect is compared with the driving condition prediction model based on the traditional BP neural network to verify the accuracy of the driving condition prediction model based on the GRU neural network.Then,the driving cycle is divided into segments and the energy consumption characteristic parameters of each segment are extracted.The principal component analysis method is used to select the characteristic parameters for energy consumption prediction.An energy consumption prediction model based on Bagging ensemble algorithm is established.and the training set and test set data are used to train and test the model respectively.The driving energy consumption prediction process based on driving cycle prediction is designed.Finally,the driving range estimation process based on battery state and driving energy consumption prediction is designed.Three electric vehicle operation scenarios under different battery states are selected to verify and analyze the driving range estimation method in this thesis.The results show that the estimation error of electric vehicle driving range is less than 10%in different scenarios,and the calculation accuracy is significantly improved compared with the industry standard.It is proved that the estimation method of electric vehicle driving range based on battery state and driving energy consumption prediction proposed in this thesis is feasible and effective.Through the research on the estimation method of electric vehicle driving range,the problem that the driving range of electric vehicle is difficult to predict can be improved,and the basis for the development of electric vehicle energy management strategy can be provided.