부산대학교 도서관

First-generation electric vehicle (EV) batteries are now retiring from their first life with 70–80%of their initial capacity and are becoming available in the market. To harness the remaining capacity of retired batteries before they reach the end of their life, researchers have proposed various methods, including backup and emergency power supplies for homes, grid-tied stationary energy storage, and storage for renewable energy applications. These applications are collectively termed second-life applications. The first step in utilizing these retired batteries is the selection of healthy cells, as cells in an EV battery pack do not degrade evenly. State-of-the-art cell selection and sorting techniques are either ineffective for industrial-grade applications or highly time-consuming processes. Therefore, this paper proposes a charging voltage relaxation pattern recognition method to efficiently sort healthy cells, powered by data-driven machine learning algorithms. A wide range of battery cycling data collected under different ambient and charging conditions is used for training, testing, and validating the proposed sorting strategy. Furthermore, a comparative analysis is conducted to demonstrate the effectiveness of the proposed strategy compared to state-of-the-art methods.