Research On Electric Vehicle Power Battery Modeling And SOC Estimation

With the rapid development of contemporary social economy and the increase in the number of cars,environmental pollution and energy shortage are becoming more and more serious.Electric vehicles have become the main direction of the development of the automobile industry by virtue of their advantages such as low noise and zero emission.As the core component of the vehicle,power battery plays a decisive role in vehicle performance such as endurance,acceleration time and safety factor.Accurate estimation of the State of Charge(SOC)of power battery can improve the efficiency and safety of the battery,and enable drivers to know the remaining driving range of electric vehicles in real time.The Battery Management System(BMS)manages batteries.The key to the accurate estimation of SOC depends on the accuracy of the model established by the power battery and the accuracy of the SOC estimation method.Therefore,the establishment of a battery model that can accurately describe the dynamic and static characteristics of the battery,with low complexity and easy engineering implementation,and the study of reliable and high precision SOC estimation method are the research focus and hotspot of improving the SOC estimation accuracy.This paper takes lithium iron phosphate power batteries commonly used in electric vehicles as the research object,focusing on power battery modeling and SOC estimation algorithm.A variable order equivalent circuit model was established to ensure the accuracy of the model and solve the contradiction between the accuracy and complexity of the battery model.The L-M method was used to optimize the iterative Extended Kalman Filter(EKF)algorithm to estimate SOC.The feasibility and effectiveness of the method were proved by simulation and experiments.Specific research contents are as follows:Firstly,the research status of battery model and SOC estimation algorithm is summarized and analyzed.Based on the analysis of the working mechanism of lithium iron phosphate battery,the influencing factors of SOC estimation were summarized,and the basic working characteristics of power battery were studied.Secondly,a variable order equivalent circuit model was established for the characteristic of "steep at both ends and flat in the middle" of the open-circuit voltage characteristic curve.The second-order RC equivalent circuit model was used in the platform region where the battery voltage changed slowly to reduce the complexity of the circuit.A third-order RC equivalent circuit model is adopted in the area of battery voltage variation(voltage range at both ends of the battery near full charge and loss of power).The Akaike Information Criterion(AIC)and least square method were used to identify the order and parameters of the model.The accuracy of the model was verified under pulse discharge and UDDS conditions,respectively.The results show that the maximum terminal voltage error of the variable order model is 11 mv.It can reflect the operating characteristics of the battery more accurately.Thirdly,based on the variable order equivalent circuit battery model,aiming at the error problem caused by ignoring higher order terms in the processing of nonlinear systems by EKF algorithm,L-M method is adopted to optimize the iterative EKF algorithm to estimate SOC,and the algorithm is verified by simulation under constant current,pulse discharge,DST,UDDS and FUDS current conditions respectively.The simulation results show that the SOC estimation error is less than 3.5% under various working conditions,which meets the requirement of accuracy index.The proposed estimation algorithm has strong tracking ability and robustness to the state measure,and is suitable for a variety of complex road working conditions.Finally,experiments were carried out on the experimental platform of the power battery management system.The experimental results show that the actual accuracy of SOC estimation is less than 4%,which verifies the accuracy,convergence and feasibility of the established model and estimation method.