State Of Charge (SOC) Estimation Of Lithium Battery

At present,with the development of energy storage technology and electric vehicles,the development and optimization of battery management system(BMS)has become a very critical problem.As a basic algorithm of the battery management system,lithium battery SOC estimation plays a vital role in the use of energy storage technology and electrical vehicles.In recent years,many scholars at home and abroad have carried out research on SOC estimation methods.Since estimating results from a single Kalman filter algorithm and an H_∞filter algorithm are not ideal,numerous researchers have carried out further transformations based on the Kalman filter algorithm and the H_∞filter algorithm,so they have obtained extended Kalman filter algorithm and unscented Kalman filter algorithm.The proposal of this type of improved algorithm has improved to some degree the SOC estimation results of the single Kalman filter algorithm and the H_∞filter algorithm,but due to ignorance of the influence of historical data,the accuracy of estimation and the rate of convergence need further improvement.In this paper,the existing methods fail to make full use of current innovation,give full play to the advantages of algorithms,and ignore the impact of historical data,and carry out further research work.This dissertation is organized as follows:(1)Based on the analysis of the characteristics of lithium battery,this paper selects the appropriate model and conducts charging and discharging experiments on the lithium battery under suitable experimental conditions.At same time,this paper uses the pulse discharge experiment method to identify the parameters of the lithium battery,It provides a good premise for SOC estimation of lithium battery.(2)When the classical H_∞algorithm(HIF)is used to estimate the state of charge(SOC)of a lithium battery,the influence of historical data is often ignored,resulting in an increase in the estimation error.In order to improve the accuracy of SOC estimation,this paper proposes an extended exponential weighted moving average H_∞algorithm(EE-HIF)for the impact of historical data.By designing the Gaussian function,the weighted distribution of data at various points can effectively reduce the estimation error caused by the inaccuracy of the lithium battery model.In addition,the superiority of the proposed method is further verified when the system model error increases.It was verified by simulation that Compared with HIF filtering algorithm and exponentially weighted moving average H_∞algorithm(EWMA),the EE-HIF algorithm offers clear advantages in terms of estimation accuracy,convergence rate and robustness.(3)Since the lithium battery system contains time-varying parameters and noise with unknown statistical characteristics,it is difficult for the existing SOC estimation methods to balance the estimation accuracy and robustness.This paper combines the benefits of the Kalman Filtering Algorithm(KF)with greater pre-stimulation accuracy and the best robustness of the H_∞Algorithm(HIF),and proposes a hybrid Kalman/H_∞(MI-KFH)filtering algorithm based on multiple innovations,Fully consider the current new information and historical information,select appropriate weights to achieve joint estimation,and effectively improve the accuracy and robustness of SOC estimation.It was verified by simulation that when the system model has time-varying parameters and noise with unknown statistical characteristics,MI-KFH combines the advantages of the Kalman filter algorithm based on multiple innovations(MI-KF)and the H_∞filtering algorithm based on multiple innovations(MI-KFH).It has the advantages of higher estimation accuracy and better robustness.