Research On SOC-SOH Joint Estimation Of On-board Lithium-ion Battery

With the rapid development of industrial level in various countries,resource shortage and environmental pollution and other problems are becoming more and more serious.Vigorously developing new energy vehicles with battery as power source has become the development direction of the current automobile industry.State of Charge(SOC)and State of Health(SOH)are two important indicators of battery management system.At present,many scholars at home and abroad have done more researches on SOC estimation method,but relatively fewer researches on SOH estimation,and the estimation accuracy of SOC and SOH is not high.To solve these problems,this paper conducts a joint estimation research on battery SOC and SOH,and the specific research contents are as follows:(1)Characteristics of lithium-ion battery.Experimental test platform of lithium-ion battery was established.The monomer type 18650lithium-ion battery was selected to be the test target.Test scheme under different discharge rates,temperatures,and working conditions was determined.Relevant impact factors and data were obtained,to support the subsequent battery modeling and SOC-SOH joint estimate.(2)Lithium-ion battery parameters identification.Considering the accuracy of the battery model and the amount of calculation,the second-order RC model is selected as the battery model.Matlab/Parameter Estimation Toolbox was used to identify battery parameters offline at 15℃,25℃,35℃ and 45℃,and polynomial fitting method was used to establish functional expressions of battery parameters with SOC and temperature influencing factors.The results show that the error between real voltage and simulated voltage at different temperatures is small.The internal parameters of lithium-ion battery change all the time,but offline identification cannot be conducted by real-time.The forgetting factor under forgetting factor in online least squares(FFRLS)identification method is usually determined by trial and error method,which seems to be less scientific and precise.Therefore,this paper proposes the least square method with time-varying forgetting factor(TVFFRLS),which can realize adaptive updating of forgetting factor.so as to reduce the error caused by the uncertainty of forgetting factor.The algorithm has good accuracy through the verification under different working conditions.(3)Joint estimation of SOC-SOH of lithium-ion battery based on ADEKF.The improved Sage-Husa algorithm was used to adaptively modify the noise,and the battery SOC was estimated based on the adaptive extended Kalman filter(AEKF)algorithm at different temperatures.The estimation accuracy of battery SOC under different temperatures could be controlled within 3.5% by cyclic pulse characteristic working conditions.The adaptive double extended Kalman filter(ADEKF)algorithm is used to estimate SOC and SOH jointly,and the SOH estimation error can be controlled within 5% through the verification of different working conditions.(4)Joint estimation of SOC-SOH of lithium ion battery based on ACKF-AEKF.As EKF for nonlinear system is linearized by first order Taylor expansion process,the second order and the above items are omitted,which will cause certain calculation error.Therefore,Unscented Kalman filter(UKF),Cubature Kalman filter(CKF),and adaptive Cubature Kalman filter(ACKF)were introduced.Through the comparison and analysis,it is found that ACKF could provide highest precision for SOC estimation accuracy,and the maximum SOC estimation error under BJDST and DST conditions is 1.05% and 0.45%,respectively.As accurate SOC is the premise of SOH estimation,ACKF-AEKF is used to jointly estimate SOC and SOH of the battery.The results show that this algorithm can not only improve the accuracy of SOC estimation,but also improve the accuracy of SOH estimation.The maximum error of SOH estimation under BJDST and DST conditions is 3.13% and 3.04%,respectively.The accuracy of SOH estimation is obviously higher than that of ADEKF algorithm.