Energy Efficient Battery Optimization Model (EE-BOM) using Machine Learning Algorithms and Harris Hawks Optimization

Electric vehicles (EVs) are gaining popularity because of their cheap running costs and positive environmental impacts. However, EVs' limited battery life is one of their biggest drawbacks. The Energy Efficient Battery Optimisation Model (EE-BOM), a unique model for early battery life detection, is presented in this work. This study makes use of a dataset from the Hawaii Natural Energy Institute that includes 14 distinct batteries that were put through more than 1000 cycles in a controlled environment. A multi-step approach is used, with feature selection coming after data collection and preprocessing with data normalisation. Additionally, for early RUL prediction, the XGBoost Approach, which combines Harris Hawk Optimisation (HHO) with Artificial Neural Networks (ANN), is used. Finding important factors affecting battery health and longevity is made easier with the help of feature importance analysis. Outlier reduction improves model accuracy, and statistical analyses show no missing or redundant data. Notably, with almost flawless predictions, XGBoost proved to be the most successful algorithm. This study emphasises how important RUL prediction is for improving battery lifetime management, especially in applications like electric cars, guaranteeing the best possible use of resources, economic viability, and environmental sustainability.