Lithium-ion battery safety is the single most crucial issue in the reliability of electric vehicles. The early state of health estimation and life prediction ensure the safe operation and timely detection of battery faults. This study aims to use voltage, capacity degradation, and EIS technique for predicting the cell’s End-of-Life. Firstly, my method can accurately predict a cell’s remaining useful life at any point in its lifetime, starting from the 10th cycle for short life cells and the 40th cycle for long life cells. Secondly, I have made significant advancements in predicting remaining useful life through EIS data by generating new features and studying the degradation pattern with aging. My results improve upon the existing literature in both the techniques.

I have worked on two datasets:

Stanford Capacity Degradation Dataset (https://data.matr.io/1/projects/5c48dd2bc625d700019f3204) Cambridge EIS Dataset (https://zenodo.org/record/3633835#.X_GO3VkzZkg)