Abstract: Battery management systems (BMS) play a vital role in the safety, efficiency, and longevity of electric vehicles (EVs). As electric mobility increases, the BMS has a critical impact on enhancing the overall vehicle perfor-mance and energy management, making its optimization vital. In this regard, this paper presents how advanced artifi-cial intelligence (AI) algorithms involving machine learning (ML), deep learning (DL), and reinforcement learning (RL) can overcome the major issues confronting BMS: precise state-of-charge (SOC) estimation, state-of-health (SOH) prediction, thermal management, and charge-discharge efficiency. AI approaches such as XGBoost and CatBoost achieve high performance for SOC and SOH predictions, with metrics like MAE, RMSE, and R² reaching values of 2.243, 3.2, 0.99, and 17.1, 23.97, 0.99, respectively, showcasing the potential for superior accuracy and robustness. The integration of AI systems facilitates improved adaptability and intelligent energy distribution, propelling the jour-ney toward a sustainable and efficient electric vehicle landscape.

Keywords: BMS, optimization EV , AI, ML, SOC, XGBoost

| DOI: 10.17148/IARJSET.2025.12216