부분 방전 곡선을 활용한 Adaptive Neuro-Fuzzy Inference System 기반 사용 후 배터리 안전성 평가 및 상태 분류 알고리즘 개발

초록

This study proposes an Adaptive Neuro-Fuzzy Inference System (ANFIS)-based classification algorithm by utilizing partial discharge curves to establish an economical and safe recycling system for used batteries. Three input variables are extracted from the partial discharge curves to construct the classification algorithm, and fuzzy logic-based membership functions and rules are designed. The ANFIS model is then trained by learning the correlation between the input variables and the safety values derived from the designed fuzzy logic system and optimizing the membership functions and rules. As a result, a classification algorithm is developed to evaluate battery safety regardless of the state of charge (SOC) and contribute to the rapid classification of used batteries for safe reuse and recycling.

키워드