Research On SOC Estimation Of Power Battery Based On Optimization Iterative Center Difference Kalman Filter

In recent years,electric vehicles have attracted the attention of an increasing number of countries,as representative of the energy conservation and environmental protection.Power battery is the power source of electric vehicle;its development directly restricts the development of electric vehicles.Therefore,more and more car manufacturers and scholars focused on the research and development of BMS.In the BMS SOC represents the mileage of the car;it is a parameter that is very important but difficult to estimate.How to accurately estimate theSOC of power battery is the key and difficult point in the filed of the battery,and even in the whole electric vehicle area.In view of the present situation,this paperfocuses on the research ofSOC estimationof power Li-ion batteries used on vehicle.The author builds a two order RCbattery model,and designs the corresponding SOC estimation algorithm,then ensures the precision and stability of the power Li-ion batteries SOC estimation.In addition,in order to meet the requirements of engineering application,the floating point arithmetic model is converted to a fixed point model.Firstly,starting fromthe SOC performance ofpower Li-ion batteries;on the basis of relevant experiments,the paper analyzes the influence of temperature,charge and discharge rate on the performance of battery SOC.Thenatwo order RC equivalent circuit modelis built combined with the actual demand,HPPC cycle testis carried out on multiple SOC points under differenttemperatureand charge\discharge rate,and the parameters of the battery is fit based on the result of cycle tests.Thepaper builds the battery model andsimulate in MTLAB,the accuracy of the battery model is verified by simulation.Secondly,based on the two order RC battery model,the paper uses EKF algorithm and CDKF algorithm to estimate the battery model's SOC.By comparing two aspects of theoretical analysis and simulation results,the paper proves that the estimation effect of CDKF algorithm for SOC is better than EKF.In addition,In order to further reduce the linearization error of CDKF algorithm,optimization ICDKF algorithm is proposed,the idea of iterative is added in the observation information updating process of CDKF algorithm,and the Levenberg-Marquardtoptimization is applied in the iteration.As a result,accuracy and stability of the estimation algorithm is improved greatly.Thirdly,in order to improve the operating efficiency of the algorithm in the single chip microcomputer,cost savings in Engineering,the paper operates on a fixed-point process on optimization ICDKF algorithmmodel of the SOC estimation,puts forward the fixed point processing strategy,and builds the algorithm modelin the Stateflow software,validates the SOC estimation results of the fixed-point model.Finally,the bench verification is carried out.The fixed-point model of optimization ICDKFalgorithm is generated to C code through Simulink/RTW-EC,and then the code is ported into battery management system.According to the SOC test rules in the national standard QC/T897-2011 and NEDC cycle working conditions,the paper test estimationaccuracy of the SOCalgorithm on the bench.