Health diagnosis and recuperation of aged Li-ion batteries with data analytics and equivalent circuit modeling

Abstract

Summary: Battery health assessment and recuperation play crucial roles in the utilization of second-life Li-ion batteries. However, due to ambiguous aging mechanisms, it is challenging to estimate battery health and devise an effective strategy for cell rejuvenation. This paper presents aging and reconditioning experiments of 62 commercial lithium iron phosphate cells, which allow us to use machine learning models to predict cycle life and identify important indicators of recoverable capacity. An average test error of 16.84% ± 1.87% (mean absolute percentage error) for cycle life prediction is achieved by gradient boosting regressor. Some of the recoverable lost capacity is found to be attributed to the non-uniformity in electrodes. An experimentally validated equivalent circuit model is built to demonstrate how such non-uniformity can be accumulated, and how it can give rise to recoverable capacity loss. Furthermore, Shapley additive explanations (SHAP) analysis also reveals that battery operation history significantly affects the capacity recovery.

Keywords

• Electrochemical energy storage
• Machine learning
• Materials science;