State Of Charge Estimation Of Power Batteries Based On Fading Kalman Filter

In recent years,the electric vehicle industry has developed rapidly in the world,and the advantages and disadvantages of one of the key factors affecting its development.Power battery management system is the core and key to ensure the efficient and safe operation of the whole vehicle,while the state of charge(SOC)estimation of power battery is the core of the power battery management system,which plays an important role in optimizing the energy management of the whole vehicle and preventing overcharge and over-discharge.As a result,it is of great theoretical significance and engineering value to study the accurate estimation of power battery SOC in depth.Based on the research object of 18650 ternary lithium battery,the purpose of this paper is to accurately estimate the SOC of power battery in actual operation condition.From the battery model and estimation algorithm,the SOC estimation of power battery is studied.Specific studies are as follows.1.The improved dual-polarization(DP)equivalent circuit model is established,and the parameters of the circuit model are identified by recursive least square(RLS)method on-line identification and off-line identification,and the adaptive output model is established.Based on the dynamic working condition experiment,the model accuracy is analyzed to confirm that the model has higher accuracy than the ordinary bipolarization model and can accurately simulate the dynamic effect of the battery.2.Because the extended Kalman filter(EKF)needs to transform the nonlinear system into a linear system by Taylor expansion when solving the nonlinear system problem,it has a certain influence on the accuracy.To solve this problem,the adaptive fading extended Kalman filter(AFEKF)is used to study the estimation of SOC,and the EKF error covariance is weighted by introducing attenuation factor to reduce the influence of the old measurement value on the estimation,and to strengthen the correction effect of the new measurement data on the filter,to improve the tracking speed and estimation accuracy.3.The MATLAB/Simulink simulation model is established,and the simulation experiments are carried out under the UDDS condition,the ECE condition and the typical DST dynamic condition respectively.The simulation results show that AFEKF can control the estimation error within 0.2% compared with EKF.The simulation results show that AFEKF can control the estimation error within 0.2% compared with EKF.35 pictures,5 tables and 57 references.