基于FORUKF-UKF的锂电池SOC联合估计研究

Research on the Joint Estimation of Lithium Battery SOC Based on FORUKF-UKF

目的 针对传统卡尔曼滤波算法估算锂电池的荷电状态(SOC),其值用R_(SOC)准确度不足的问题,提出一种分数阶鲁棒无迹卡尔曼滤波联合无迹卡尔曼滤波(FORUKF-UKF)方法估计锂电池SOC。方法 在动态应力测试(DST)下,采用自适应遗传算法(AGA)对锂电池分数阶模型(FOM)进行参数辨识;在FOM的基础上将无迹变换(UT)技术与H∞观测器结合提出FORUKF算法,并与UKF联合实现SOC估计;联合估计器中的UKF实时估计电池模型中的欧姆电阻R_(0),并反馈至FORUKF算法中估算得到SOC;最后在北京动态应力测试(BJDST)下与拓展卡尔曼滤波(EKF)、分数阶无迹卡尔曼滤波(FOUKF)进行比较验证。结果 在估计SOC的过程中FORUKF-UKF方法相对于EKF、FOUKF和FORUKF始终保持了最高的估计精度,展现了更好的鲁棒性。结论 FORUKF-UKF方法在估计锂电池SOC方面比EKF、FOUKF和FORUKF算法具备更好的准确性和鲁棒性。

Objective In order to address the issue of inadequate accuracy in estimating the state of charge(SOC) of lithium batteries using traditional Kalman filtering algorithms,this study proposes a fractional order robust unscented Kalman filter-based unscented Kalman filter(FORUKF-UKF) method for SOC estimation.The estimated value of the lithium battery's state of charge is denoted by R_(SOC).Methods An adaptive genetic algorithm(AGA) was employed to identify the parameters of a fractional order model(FOM) of the lithium battery during dynamic stress testing(DST).The FORUKF algorithm was proposed by combining the unscented transform(UT) technique with the H∞ observer based on FOM,and the SOC estimation was jointly realized with the UKF.The UKF in the joint estimator realized real-time estimation of the Ohmic resistance R_(0) in the battery model and fed it back to the FORUKF algorithm to estimate SOC.Finally,comparisons and verifications were conducted with extended Kalman filtering(EKF) and fractional order unscented Kalman filtering(FOUKF) using Beijing dynamic stress testing(BJDST).Results The results showed that the FORUKF-UKF method consistently achieved the highest estimation accuracy in the SOC estimation process compared with EKF,FOUKF,and FORUKF,demonstrating better robustness.Conclusion The FORUKF-UKF method has better accuracy and robustness than the EKF,FOUKF,and FORUKF algorithms in estimating the SOC of lithium batteries.