AI-BASED PREDICTIVE THERMAL MANAGEMENT OF ELECTRIC VEHICLE BATTERIES FOR IMPROVED PERFORMANCE, SAFETY AND BATTERY LIFE

Abstract

The rapid adoption of Electric Vehicles (EVs) has increased the demand for safe, efficient, reliable and long-life lithium-ion battery systems. One of the major challenges associated with EV batteries is thermal management. During charging and discharging, electrochemical reactions and internal resistance generate heat. Excessive temperature and temperature non-uniformity can negatively affect battery performance, accelerate degradation and increase safety risks. Conventional Battery Thermal Management Systems (BTMS) generally use fixed or rule-based control strategies. These approaches may not respond optimally to rapidly changing operating conditions such as load current, state of charge, ambient temperature and driving conditions. Artificial Intelligence (AI) and Machine Learning (ML) provide an opportunity to predict battery temperature and adapt the cooling requirement according to real-time conditions. This research proposes an AI-based predictive thermal management framework for lithium-ion batteries used in electric vehicles. Battery operating parameters such as current, voltage, state of charge, ambient temperature and cooling conditions are considered as input variables. Machine-learning models such as Artificial Neural Network (ANN), Random Forest (RF) and XGBoost can be trained to predict battery temperature. The predicted temperature can subsequently be used to determine an appropriate cooling requirement. Model performance can be evaluated using Mean Absolute Error (MAE), Root Mean Square Error (RMSE) and coefficient of determination (R²). The proposed approach aims to maintain battery temperature within a suitable operating region while reducing unnecessary cooling energy consumption.