Research On Modeling And State Of Charge Estimation Of Lithium Ion Battery

Lithium-ion batteries,as the power source of electric vehicles,which can effectively solve the problems of energy consumption and environmental pollution,have been widely used in electric vehicles.A stable battery management system can effectively control the normal operation of the battery pack.State of Charge(SOC)estimation is one of the important modules.Accurately estimating lithium ion battery state of charge of can improve the safety performance of the battery,avoid overcharging and overdischarging,extend useful life,and improve battery utilization.This paper focuses on the research of battery model and state-of-charge estimation.Taking lithium iron phosphate battery as the research object,partnership for a new generation of vehicles battery model has been established in the time domain and a factional order model in the frequency domain has been determined with using the modeling method of fractional components and analysis of dynamic characteristics of electrochemical impedance spectroscopy,respectively.And combined with open circuit voltage experiments,hybrid pulse characteristic test(HPPC)and Beijing Bus Dynamic Stress Test(BBDST),state of charge of lithium battery has been estimated based on extended Kalman filter algorithm.Firstly,the paper outlines research background,the future prospects for the development of electric vehicles and the state of charge estimation module in the battery management system.Lithium iron phosphate batteries were identified as the research object through analysis of two types of lithium batteries.And the current status of domestic and foreign research on the equivalent circuit model and the method of estimating the state of charge of the lithium battery are analyzed in detail.Secondly,PNGV equivalent circuit model in time domain and in frequency domain are established respectively.When establishing PNGV model,parameters in the model are obtained by offline identification through open circuit voltage experiment and hybrid pulse characteristic test.Then verify the accuracy of the model at different SOCs.The results show that the maximum relative error of the PNGV model is 2.88%.Analysis of dynamic characteristics of electrochemical impedance spectroscopy shows that SOC has a greater influence on the mid frequency region.Determine the fractional model by fractional modeling and fitting the fractional components in frequency domain.By fitting impedance curves in thefrequency domain and based on fractional calculus(G-L)definition in the time domain,system differential equation is solved and immune hybrid particle swarm optimization algorithm is used to identify model fractional component parameters.In Beijing Bus Dynamic Stress Test operation,error,the root-mean-square error,average absolute value error and average relative error of the model are 0.02 V,0.66%,0.53% and 0.16%,respectively.Finally,by discretizing system equations of state and observation equations,model parameters are substituted to design extended Kalman filter.Extended Kalman Filter(EKF)is used to estimate battery state of charge in BBDST operation based on established models.The maximum relative error of fractional model does not exceed 1%,and the maximum relative error of PNGV model is 2.78%.The results show that the accuracy of SOC estimation by fractional order modeling is higher than that of PNGV model,which improve the accuracy of SOC estimation.The extended Kalman filter algorithm has high accuracy and small error,and can be used as a practical method for estimating SOC.