Estimation of Model Parameters of Lithium Batteries Based on Kalman Filtering Optimized by Dung Beetle Algorithm

The identification of parameters for lithium batteries is an important basis for battery state prediction. An improved kalman filtering (KF) based on dung beetle optimizer (DBO) is proposed for online identification of battery model parameters. This method utilizes the rapid global search for optimal solutions characteristic of DBO to optimize the covariance matrices of process noise and observation noise in KF, thereby improving the accuracy of identifying battery model parameters. Simulation experiment data shows that compared to the parameter identification results based on unoptimized KF, the variance that the identification results of this method compared to the true values are significantly reduced, resulting in the predicted parameter values are closer to the true values.