Early Diagnosis of Accelerated Aging for Lithium-Ion Batteries With an Integrated Framework of Aging Mechanisms and Data-Driven Methods

Accelerated aging is a significant issue for various lithium-ion battery applications, such as electric vehicles, energy storage, and electronic devices. Effective early diagnosis is prominent to restrict battery failure. Typical battery classification data-driven methods are structured to capture features from data without considering the underlying aging mechanism. On the other hand, analysis of the detailed aging mechanism that can generate electrochemistry-based models can be highly complicated and may not be suitable for real-time battery management. In this article, the accelerated aging diagnosis method is systematically investigated. The accelerated aging mechanisms of the Li[NiCoMn]O2 (NCM) battery are analyzed by the nondestructive quantitative diagnostic method. We prove the feasibility of accelerated aging diagnosis based on the accelerated aging mechanism analysis. An integrated framework of aging mechanisms and data-driven methods (IFAMDM) is introduced for lithium-ion battery-accelerated aging diagnosis. Highly adaptable features reflecting the accelerated aging mechanism are proposed for lithium-ion battery-accelerated aging. Then, we propose a combination method to diagnose battery-accelerated aging. The IFAMDM was verified on two types of battery datasets. The IFAMDM is proved to be highly generic and accurate for lithium-ion battery-accelerated aging diagnosis at the 100th cycle.