Accurate Prediction of Electric Vehicle Battery Lifespan Using XGboost and Ensemble Methods

Thanks to their low running costs and positive environmental impacts, electric cars, or EVs, are growing in popularity. Still, the limited battery life of EVs is one of the primary drawbacks. The present work offers a thorough method for estimating the Nickel-Manganese Cobalt-Lithium Cobalt Oxide (NMC-LCO) batteries' Remaining Useful Life (RUL). Selection of features and outlier removal are the next steps in the multi-step process that begins with data gathering and preprocessing. The development of the RUL prediction model makes use of ML methods such as XG-Boost, Bagging-Regressor, Light-GBM and Cat-Boost. Critical factors affecting battery longevity and health may be found with the use of feature importance analysis. Analytical analyses show that there are no missing or duplicate data points, and the elimination of outliers improves model accuracy. XGBoost stood up as the most efficient algorithm, yielding almost flawless predictions. In applications like as electric cars, where resource optimisation, cost effectiveness, and environmental sustainability are critical, this study emphasises the value of RUL prediction in improving battery lifetime management.