Research On State Of Charge And Remaining Useful Life Prediction For Vehicle Power Battery

Lithium-ion batteries(LIBs)have become the preferred energy supplier for electric vehicle power storage systems due to their high energy density,low cost,and excellent safety performance.However,with the growing demand for electric vehicle safety,the online monitoring and evaluation of the operating status and safety performance of LIBs have become a crucial task in the development of efficient battery management systems.In response to practical application requirements,this paper conducts research on predicting the state of charge(SOC)and remaining useful life(RUL)of lithium-ion batteries.The research work mainly includes the following aspects:(1)Considering that the parameters of a battery are not fixed and vary with changes in its SOC,this paper proposes a SOC-dependent second-order RC equivalent circuit model.Based on this model,an exponential fitting method and a least squares method with a forgetting factor are employed to identify the battery parameters,and the trend of each parameter with respect to SOC is analyzed using a fifth-order polynomial function fitting.Finally,the proposed SOC-dependent second-order RC equivalent circuit model is implemented in MATLAB/Simulink to verify its accuracy.(2)On the basis of considering the influence of SOC variation,the second-order RC equivalent model is applied to achieve real-time and accurate estimation of battery SOC by utilizing particle filter(PF)algorithm.The accuracy of the proposed model is validated in constant current and dynamic working conditions,by comparing with the ordinary second-order RC model.Additionally,different filtering algorithms are employed to estimate SOC,demonstrating the universality of the proposed model in SOC estimation.(3)In view of the highly nonlinear relationship between the Open Circuit Voltage(OCV)and SOC in batteries,Back Propagation Neural Network is employed to fit the OCV-SOC relationship with high accuracy.Additionally,an extended particle filter algorithm is proposed by combining extended Kalman filter with particle filter algorithm to estimate SOC.The accuracy and robustness of the proposed algorithm are verified by simulation experiments.(4)A precise and stable residual life prediction method for lithium-ion batteries is proposed based on a dual-exponential empirical model and combined particle swarm optimization algorithm with PF algorithm.Subsequently,the accuracy of the proposed method is effectively validated through two lithium-ion battery prediction experiments.Furthermore,based on the work of this paper,a comprehensive performance prediction and analysis software for power batteries in electric vehicles is designed,which is used to predict the SOC and RUL of lithiumion batteries.