Research On Fault Prediction Of Electric Vehicle Power Battery Based On Data Mining

With the vigorous development of the new energy industry of electric vehicles,energy problems and environmental problems have been greatly alleviated.However,due to the technical reserve problem,many self-ignition phenomena of electric vehicles have appeared at home and abroad,and the safety problems have also attracted people's attention.One of the important factors in the development of electric vehicles,it is of great practical significance to predict and diagnose the failure of batteries.In this era,data has become a breakthrough point in technological change.Whoever can better dig out the hidden information in the data can get a head start in the market.In this paper,the battery pack data collected by the "National New Energy Vehicle Monitoring Platform" is used to study the real-time fault of the power battery.Based on the real-time driving data of the electric vehicle,the fault diagnosis of the battery is realized by means of big data.prediction.The research content of this paper is as follows:1.Using PCA dimension reduction and K-means cluster preprocessing to process and analyze the cell voltage data,identify and diagnose various faults of cell;2.Taking the voltage of the electric vehicle battery pack as the learning sample,using the basic principle of the least squares support vector machine regression algorithm(LS-SVR),the fault measurement model based on the total voltage of the battery pack is established,and the total voltage is realized.Realized prediction of total voltage overvoltage and undervoltage faults;3.In order to improve the prediction accuracy of the model,the grid method optimization and K-fold cross-validation method are used to optimize the parameters of the prediction model,and then the model is trained by using the cell voltage data to predict the state trend and fault prediction.The comparison between the support vector machine and the least squares regression machine improves the prediction accuracy and can diagnose and predict the faults generated immediately by the electric vehicle power battery.4.Pre-processing various data of battery packs of long-running power batteries,establishing feature vectors in daily units and scale-changed feature vectors in weeks,mining data fault features,and establishing based on Arrhenius equation The battery's long-term performance degradation prediction model.The real-time data is brought into the prediction model for verification,and the prediction effect is better.