A P-Value Approach for Real-Time Identifying the Capacity Regeneration Phenomenon of Lithium-ion Batteries

Lithium-ion batteries (LIBs) are favored as the primary energy storage option as a result of their high energy density, extended cycle life, and low self-discharge rate. However, the health condition of LIBs is critical for ensuring the overall safety, reliability, and optimal performance of electric vehicles (EV s) and other energy storage systems. The capacity regeneration phenomenon (CRP) causes an increase in the available capacity of LIBs during random periods of inactivity caused by variability in human behavior, such as the driving habits of EVs. The presence of CRP can have a significant impact on the accuracy of the state-of-health (SOH) estimation and the remaining useful life (RUL) prediction that most researchers often neglect, resulting in flawed SOH estimation, premature replacement, or unexpected battery failures. Therefore, developing accurate and reliable SOH estimation algorithms that consider various operating conditions and degradation mechanisms is crucial to ensure the optimal performance, reliability, and longevity of LIBs. This paper proposes a statistical approach using hypothesis testing and a test statistic with corresponding p-value that can effectively identify CRP in real time with 99.4 % accuracy.