State of health estimation of individual batteries through incremental curve analysis under parameter uncertainty

Abstract

Abstract The assessment of State of Health (SOH) plays a decisive role in diagnosing the health condition of Lithium‐Ion Batteries (LIBs). However, SOH estimation, particularly for individual battery cells, remains underexplored, especially under working conditions and aging patterns where battery parameters cannot be fully determined. This research conducted a comparative analysis of the parameter sensitivity among three methods and proposed a novel approach to estimate the SOH in large‐capacity batteries. The proposed method integrates multi‐feature extraction with artificial intelligence techniques. Specifically, various Health Index sets (HIs) reflecting Incremental Capacity morphological features are extracted from the charging curves of LIBs. Subsequently, a method is proposed to fuse these HIs using an artificial neural network to achieve precise SOH estimation. The effectiveness of the proposed method is validated through extensive long‐term degradation experiments on Lithium Cobalt Oxide batteries. The results confirm significant attributes of the method, including high estimation accuracy, reliability, and robustness against small‐scale inconsistencies.

Keywords

• battery management systems
• energy storage