Parameter identification of lithium-ion battery based on least squares algorithm with variable forgetting factor

Abstract

Power battery performance plays a pivotal role in the comprehensive performance of electric vehicles, and thus accurate identification of the parameters of the lithium-ion battery model is crucial for subsequent state-of-charge estimation and state-of-health prediction of the battery system. In order to improve the accuracy of parameter identification algorithm of lithium-ion battery model, a second-order RC equivalent circuit model of the battery is established, and a least-squares algorithm based on variable forgetting factor is used to identify the parameters of the lithium-ion battery model online. By building a test platform for charge and discharge experiments, based on the experimental data of two different operating conditions, the proposed algorithm, recursive least squares algorithm and traditional forgetting factor least squares algorithm are used to identify the parameters, and the accuracy of the proposed algorithm is described based on the comparison of the error between the estimated port voltage and the actual value obtained from the experimental test. The experimental results show that the recursive least squares algorithm based on the variable forgetting factor shows fast convergence and high estimation accuracy in the identification of lithium-ion battery parameters.

Keywords

• lithium-ion battery
• model parameters
• online identification
• variable forgetting factor
• second-order rc
• least squares algorithm