A Novel Approach for Identification of Lithium-ion Battery Equivalent Circuit Model Parameters Based on Multi-dimensional Adaptive Feedback Regulation

Le résumé fourni par la source

The global optimal battery model parameter is essential for accurate state estimation of battery packs in electric vehicles. It occurs that identification accuracy is low, computational load is high, and computation falls into local optimum when the traditional battery model parameter identification methods are used. Therefore, a multi-dimensional adaptive feedback regulation (MAFR) algorithm for parameter identification is proposed based on the simulation model of Simulink in this paper. The rationale of the proposed method is selecting the multi-dimensional evaluation index that takes into account overall and local errors, and constructing a correlation function as a criterion between model parameters and evaluation indexes. Then, the adaptive step size is used by the algorithm to continuously adjust the model parameters. The identified parameters can be obtained when the multi-dimensional evaluation index is of optimal value. The experimental results show that the proposed method can efficiently and accurately converge to the global optimal model parameter (GOMP). The root mean square error (RMSE) of the identified model can reduce to 29 mV, with the mean absolute error (MEA) being less than 16.7 mV. Besides, the accuracy and efficiency of SOC estimation based on GOMP are improved, and at the same time the accuracy and validity of the proposed method are verified.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.