Improved covariance matching-electrical equivalent modeling for accurate internal state characterization of packing lithium-ion batteries

As for the cell-to-cell inconsistency of packing lithium-ion batteries, accurate equivalent modeling plays a significant role in the working characteristic monitoring and improving the safety protection quality under complex working conditions. In this work, a novel covariance matching-electrical equivalent circuit modeling method is proposed to realize the adaptive working state characterization by considering the internal reaction features, and an improved adaptive weighting factor correction-differential Kalman filtering model is constructed for the iterative calculation process. A new parameter named state of balance is introduced to describe the cell-to-cell variation mathematically by forming an effective influence correction strategy. An adaptive covariance matching method is investigated to update and transmit the noise matrix for high-power energy supply conditions, in which the weighting factor correction is conducted by considering the coupling relationship to improve the prediction accuracy. Experimental tests are conducted to verify the estimation effect, in which the closed-circuit voltage responds well corresponding to the battery state variation. The maximum closed-circuit voltage traction error is 1.80%, and the maximum SOC estimation error for packing lithium-ion batteries is 1.114% for the long-term experimental tests with the MAE value of 0.00481 and RMSE value of 5.44085E-5. The improved covariance matching-electrical equivalent circuit modeling method provides a theoretical foundation for the reliable application of lithium-ion batteries.

Automated summary

The study introduced a novel covariance matching-electrical equivalent circuit modeling method with an improved adaptive weighting factor correction and differential Kalman filtering model for characterizing the adaptive working state of lithium-ion batteries. Experimental tests showed good response of the method to battery state changes, leading to improved estimation accuracy.