Remaining Useful Life Prediction for Lithium-Ion Batteries Based on Hybrid Ensembles Allied with Data-Driven Approach

Capacity fade in lithium-ion batteries (LIBs) poses challenges for various industries. Predicting and preventing this fade is crucial, and hybrid methods for estimating remaining useful life (RUL) have become prevalent and achieved significant advancements. In this paper, we introduce a hybrid voting ensemble that combines Gradient Boosting, Random Forest, and K-Nearest Neighbors to forecast the fading capacity trend and knee point. We conducted extensive experiments using the CALCE CS2 datasets. The results indicate that our proposed approach outperforms single deep learning methods for RUL prediction and accurately identifies the knee point. Beyond prediction, this innovative method can potentially be integrated into real-world applications for broader use.