基于模糊FFRLS-IMIUKF的锂离子电池SOC估计OA

The time-varying characteristics of Li-ion battery parameters and the state of charge(SOC)estimation are susceptible to initial errors and noise interference,a collaborative estimation method combining fuzzy adaptive forgetting-factor recursive least squares(FFRLS)and improved multi-innovation unscented Kalman filter(IMIUKF)is presented.Firstly,based on first-order RC equivalent circuit model,a fuzzy controller is designed to dynamically adjust the forgetting factor of FFRLS,estimate model parameters in real time.Secondly,on the basis of the traditional multi-innovation unscented Kalman filter(MIUKF),the IMIUKF algorithm limits the innovation vector to the current and two previous steps and incorporates a posterior innovation correction mechanism to enhance tolerance to initial SOC errors and process noise.Validation under the urban dynamometer driving schedule(UDDS)driving cycle using NCR-18650GA Li-ion batteries demonstrate that,even with±20％initial SOC error,the proposed method achieves a mean absolute error(MAE)of 3.22％and a root mean square error(RMSE)of 3.16％,outperforming extended Kalman filter(EKF),unscented Kalman filter(UKF)and standard MIUKF algorithms,while maintaining real-time performance.