Research And Realization Of Life Prediction Of Ternary Lithium Ion Batteries Under Working Conditions

With the rapid development of new energy vehicles,as the main source of power for new energy vehicles,the life expectancy of lithium-ion batteries is particularly important.This paper takes a high specific energy ternary lithium ion battery with a rated capacity of 114 Ah as the research object,and carries out life modeling and life prediction software development on the battery.First,the experimental design of the battery under temperature and rate stress was carried out.Based on the experimental data,the main reasons for the decline of the ternary battery at different temperatures and different rates under normal temperature and higher temperature were analyzed.The capacity of the battery under temperature and rate stress decay.Secondly,the battery decay model under the conditions of single stress and coupled stress is established,including temperature stress life model,rate stress life model,variable temperature stress life model,variable rate stress life model and coupled temperature and rate stress life model.The particle swarm optimization algorithm is used in the problem of empirical model fitting and parameter solving,which solves the problem of multi-parameter solution under the empirical model..Thirdly,a data-driven battery life prediction model based on Gaussian process regression is established,and the ternary lithium-ion battery life prediction of a single nuclear function and a combined nuclear function is compared and analyzed.Because a single nuclear function is difficult to express complex actual working conditions,this study is The combined kernel function(Matern + RBF)is used to predict the life of the ternary lithium-ion battery.The modeling process uses a cross-validation method to test the generalization ability of the model.The battery life prediction accuracy based on the combined kernel function is within 2%,which solves the problem that the single kernel function has a large error in battery life prediction.Finally,a GUI interface was designed using Python,and a comprehensive software capable of battery life prediction was developed.The software realizes the prediction of the cycle life of ternary lithium ion batteries under different temperature and rate stress.The software interface is simple and clear,the error of battery life prediction is small and can be used across platforms,and the portability is good.