A Novel ARD-GPR Approach for Li-Ion Battery SoH Estimation Using Frequency Response Analysis

As the use of lithium-ion batteries continues to grow, a large number of these batteries are reaching the end of their first life. Making informed decisions about battery retirement, second-life applications, and fair market value requires an accurate estimation of the battery's state of health (SoH). This paper presents a data-driven method for SoH estimation using features extracted from the frequency response of the battery in the form of electrochemical impedance spectroscopy (EIS) data. A novel model based on Automatic Relevance Determination Gaussian Process Regression (ARDGPR) is proposed, identifying the most important EIS features for accurate SoH prediction. Additionally, the paper introduces the design of a hardware platform capable of real-time EIS measurement and on-board SoH estimation. The proposed approach supports reliable battery health assessment for first- and second-life applications.