A Highly Robust State of Health Estimation Method for Lithium-lon Batteries Based on ECM and SGPR
Accurately estimating the state of health(SOH)of lithium-ion batteries is of great significance in ensuring the safe operation of the battery system.Addressing the issue where traditional SOH estimation methods fail under variable working conditions,an online SOH estimation method for lithium-ion batteries based on equivalent circuit model(ECM)and sparse Gaussian process regression(SGPR)is proposed.During the constant current charging process,the parameters of the ECM of lithium-ion battery are dynamically identified by two online filters,based on which,a condition-insensitive health indicator is constructed.In combination with the SGPR,the indirect SOH estimation is achieved.This method uses the unified signal processing method and feature mapping model under various working conditions,and features strong robustness with low redundancy.The experimental results show that the average absolute error of the method proposed under various working conditions does not exceed 0.94％,and the root mean square error stays below 1.12％.When benchmarked against existing methods,this method has significant advantages in comprehensive performance.
lithium-ion batterystate of health(SOH)health indicatorparticle filtersparse Gaussian process regression(SGPR)
NETL
NSTL
万方数据
