Study On State Of Health Estimation And Remaining Useful Life Prediction Of Lithium Ion Power Battery

In order to alleviate the pressure of environmental pollution and energy shortage caused by rapid economic development,the electric vehicle industry has received great attention from the government and enterprises and achieved rapid development.At present,electric vehicles mainly use lithium-ion power batteries,but the safety of batteries is always one of the important issues in the development of electric vehicles.Current studies on the safety of lithium-ion power batteries mainly focus on two aspects: State of Health(SOH)and Remaining Useful Life(RUL).Domestic and foreign scholars have used a variety of methods to predict SOH and RUL for lithium-ion power batteries,but the prediction accuracy is limited frequently.In this paper,based on the data of lithium-ion battery from NASA Prediction Center of Excellence,the failure characteristics of lithium-ion battery were extracted,and the grey correlation analysis method and feature combination comparison were used for feature selection.The SOH and RUL of lithium-ion power battery were predicted based on the improved least square support vector machine.The main work and innovation of this paper are as follows:(1)The failure characteristic parameters of lithium-ion power battery were extracted,and the grey correlation degree of different failure characteristic parameters as the input of lithium battery SOH estimation model was calculated.Based on this,the constant current charging time,constant voltage charging capacity and constant voltage charging average temperature were determined as the input failure characteristic parameters of lithium battery SOH estimation model.(2)The data of lithium-ion battery under three charging conditions of low temperature,room temperature and high temperature were selected as the training set,and the prediction model of lithium-ion power battery SOH based on least squares support vector machine was optimized by using genetic algorithm,and the better prediction effect of lithium-ion power battery SOH was obtained.(3)Extracting constant current charging time and constant voltage charging capacity as lithium ion power battery remaining service life forecast model of input failure characteristic parameters,and using the particle swarm algorithm to optimize the lithium ion power battery remaining service life of the model parameters,obtained the good effect of lithium ion power battery remaining service life prediction,to predict remaining service life of the battery provides the reference basis.