Robust noise-correction recursive least square method for parameter identification of equivalent circuit model in battery management system using Bayes’ theorem-based preprocessing technique

The recursive least square (RLS) algorithm is considered as one of the popular methods for the parameter identification in battery management systems (BMSs) models due to its high estimation accuracy. Nevertheless, the performance of a traditional RLS algorithm is prone to be biased, when a strong interference from random noise or from a malfunction of an analog-to-digital converter occurs in the sensed input signals, causing the estimated results to deviate from the true values. To cope with this issue, this paper proposes the noise-correction RLS (NC-RLS) method using Bayes’ theorem preprocessing to enhance the robustness and estimation accuracy of the RLS method for the parameter identification of the Li-ion battery model under the disturbance of strong noises or outliers. The proposed method was implemented to identify the parameters of the equivalent circuit model (ECM) of the lithium-ion battery under two standard driving cycles, Urban Dynamometer Driving Schedule (UDDS) and California Unified Cycle (LA92), which represent the daily driving conditions. The estimation results of the proposed method are verified and compared with the reference values obtained from the MATLAB identification tools. The result analysis indicates that the proposed method has improved the robustness and the accuracy of estimation by at least 12.7% and 8.8% in UDDS and LA92, respectively, and enhanced the estimation robustness against at least 60 outliers.