A novel Bayesian multivariate linear regression model for online state-of-health estimation of Lithium-ion battery using multiple health indicators

Abstract

An accurate state of health is critically significant to evaluate the battery aging level and to ensure electric vehicle security and reliability. This paper presents a novel Bayesian-based method to quantify battery capacity degradation using multiple health indicators extracted from the battery Thevenin model at different temperatures. In this method, the linearity and monotonicity between the extracted health indicators and capacity degradation are analyzed by exploiting the correlation analysis. To estimate the battery's state of health (SOH) accurately, the Bayesian multivariate linear regression is introduced to develop an online state of health estimator. Experimental tests are conducted on two batteries with the same specifications to verify the efficiency of the proposed method. Additionally, the interval estimation of numerical results contains reasonable uncertainty expression, which validates the robustness and reliability of the proposed method.