Research On Co-Estimation Of Lithium-Ion Battery State At Multiple Time Scales Based On AFF-RLS And DAKF

With the promotion of the concepts of "carbon neutralization" and "carbon peaking",China’s new energy industry is developing rapidly,and the shipment of power lithium-ion batteries is increasing rapidly,which are widely used in electric vehicles,energy storage and other fields.State of Charge(SOC)and State of Health(SOH)are important theoretical basis for battery management system to effectively manage power lithium-ion batteries.This topic focuses on the adaptive collaborative estimation of SOC and SOH of power lithium-ion battery.(1)A series of battery characteristic experiment was designed to analyze the internal resistance,capacity,open-circuit voltage and other parameters of the battery under different charging and discharging rates and different temperatures.At the same time,cyclic charging and discharging experiment and aging characteristic analysis were carried out to prepare for the equivalent circuit modeling.(2)According to the experiment of battery performance,second-order RC equivalent circuit model is constructed,through curve fitting method to realize the offline parameter identification under different aging conditions,by forgetting factor least square algorithm to achieve the precise temperature condition on-line parameter identification,In addition,the model verification under dynamic stress test conditions of Beijing public transportation,hybrid pulse power characteristic test and dynamic stress test conditions were carried out.(3)Aiming at the joint estimation method of SOC and SOH,a dual Kalman filtering algorithm and a dual adaptive Kalman filtering algorithm are established,and a adaptive forgetting factor dual adaptive Kalman filtering algorithm is established by combining the forgetting factor least square method,which realizes the joint estimation of the state under a variety of complex working conditions,including Beijing bus dynamic stress test conditions,hybrid pulse power characteristics test conditions,and dynamic stress test conditions.Compared with other algorithms,the results show that the maximum mean absolute error of the proposed algorithm for SOC estimation is53.8750% lower than that of the dual Kalman filtering algorithm,and 57.4792% than that of the dual adaptive Kalman filtering algorithm.For SOH estimation,the maximum mean absolute error of the algorithm in this paper is 40.4522% and 41.9537% lower than that of the dual Kalman filtering algorithm and the dual adaptive Kalman filtering algorithm respectively.(4)With three sampling interval time scales of 0.1s,0.3s and 1s,the adaptive forgetting factor dual adaptive Kalman filtering algorithm proposed in this paper is used to realize the state collaborative estimation under the hybrid pulse power characteristics test conditions test condition.The results show that the prediction error increases with the increase of the sampling interval,but the algorithm in this paper still maintains a better performance at the maximum sampling interval.The average absolute error of SOC estimation is kept within 1.5%,while the average absolute error of SOH estimation is kept within 2.0%,which has obvious advantages over other algorithms.The experimental results show that the adaptive forgetting factor dual adaptive Kalman filtering collaborative estimation method proposed in this topic can effectively realize the state join estimation,and provide a theoretical basis for the effective management and safe application of power lithium-ion battery.