Study On Estimation Of SOC Of Lithium Battery By EKF Algorithm Considering Aging Correction

Considering the battery aging problem in the use of battery pack is one of the key problems to realize reliable safety monitoring.This requires a set of system to complete this task efficiently.Battery management system(BMS)came into being.State of charge(SOC)is one of the important references for battery management system to manage battery pack.The accurate estimation of SOC of lithium-ion battery considering battery aging is the key parameter for BMS to monitor the operation state of battery pack in the whole life cycle.This paper studies the SOC estimation of lithium-ion battery,mainly including:Firstly,the principle and working characteristics of lithium battery aging are analyzed,and the related lithium battery charge discharge experiments at different rates and battery accelerated aging experiments are designed.The characteristic experiment of lithium battery is carried out,the voltage capacity curve under standard current condition is measured,the battery discharge capacity experiment under different discharge rates is carried out,and the coulomb efficiency is calibrated.At the same time,the battery aging experiment is carried out to obtain the battery capacity attenuation data,which provides data support for the parameter identification of battery equivalent model and SOC estimation.Secondly,the analysis found that during the cycle charge and discharge process of the battery,the accuracy of the battery electrochemical model identified by the initial experimental data continued to decline,increasing the accuracy error of SOC estimation,while using AEKF algorithm and considering the aging data to modify the model parameters to reduce the increase of the accuracy deviation of battery SOC estimation caused by battery aging.The SOC estimation methods mainly include:1)In order to solve the problem of battery parameter drift caused by battery aging and other factors,which leads to the decline of battery model accuracy,in addition to accurately identifying the model parameters by using the least square method,it is also necessary to add links such as parameter identification based on voltage curve integral and adaptive updating of the mean and variance of noise.2)For comparison,extended Kalman filtering(EKF)and adaptive extended Kalman filtering(AEKF)are used to estimate SOC in this paper;Compared with EKF,AEKF algorithm of adaptive filtering algorithm can automatically adjust the noise variance and further reduce the problem of SOC estimation accuracy reduction due to model parameter drift caused by battery aging.The experimental results show that AEKF algorithm is better than EKF algorithm in the accuracy of battery SOC estimation.Finally,the hardware platform is built to harden the algorithm,and multi condition experiments are carried out to verify the effectiveness of the algorithm.Experiments show that the BMS designed in this paper can meet the requirements of SOC estimation accuracy of series battery pack.