An Optimal Clustering Algorithm for Second Use of Retired EV Batteries Using DBSCAN and PCA Schemes Considering Performance Deviation

This paper proposes an optimal clustering algorithm considering performance deviation of parameters and data preprocessing method for reusing retired batteries. The proposed method regroups batteries by considering the density and performance deviation of the retired battery dataset through a clustering algorithm using density-based spatial clustering of applications with noise (DBSCAN). Additionally, the performance of the algorithm was improved through data preprocessing using a principal component analysis (PCA) that prevents the computational complexity and overfitting of clustering algorithm. The feasibility of the proposed algorithm is verified by comparing with general clustering algorithms such as the k-means clustering and Gaussian mixture model.