Abstract
Battery technology stands at the forefront of sustainable energy solutions, offering clean and efficient power generation for a wide range of applications, from transportation to stationary power generation. A critical aspect of the reliability and longevity of battery systems is the accurate prediction of the remaining useful life (RUL) of battery cells and stacks. Predictive maintenance strategies aimed at forecasting RUL are essential for optimizing system performance, minimizing downtime, and reducing maintenance costs. In light of this, this paper focuses on advancing predictive maintenance methodologies for battery-powered heavy-duty vehicles (b-HDVs) using data-driven methods. The report explores the efficacy of various machine learning (ML) algorithms in predicting RUL, including XGBoost, Random Forest, Decision Tree, and Long Short-Term Memory (LSTM) networks. These algorithms are selected for their suitability in handling complex temporal data and their proven effectiveness in predictive maintenance tasks across various domains. Through rigorous comparison and evaluation, this study aims to identify the most promising approach for RUL prediction in battery packs, thereby contributing to the advancement of predictive maintenance methodologies in the battery industry.