A Precise Life Estimation Method for Retired Energy Storage Batteries Based on Energy Storage Batteries Attenuation Characteristics and XGBoost Algorithm

Abstract

With the increasing scale of energy storage batteries, the number of retired energy storage batteries is also rapidly increasing, and the energy storage life, as an important indicator for evaluating the safety of retired energy storage, has received widespread attention. The existing methods for estimating the life of retired energy storage have the problem of considering attenuation characteristics (AC) that do not fully reflect the actual operating characteristics of batteries, which can affect the estimation accuracy. Therefore, this article proposes a precise estimation method for the life of retired energy storage batteries to improve the accuracy of estimating the life of retired energy storage. Firstly, an AC extraction method was proposed to fully explore the potential relationship between various feature data and the operational status of retired energy storage. Then, a method was proposed to screen and optimize ACs that are highly correlated with the life of retired energy storage batteries, selecting ACs that are highly correlated with the life of retired energy storage batteries. Finally, an accurate method for estimating the life of retired energy storage batteries was proposed, which effectively improved the accuracy of estimating the life. The proposed method was validated on the NASA dataset with an accuracy of no less than 99.784% for estimating life.

Keywords

• Estimation of retired energy storage life
• XGBoost
• principal component analysis
• correlation coefficient