Hybrid IoT–ML-Assisted PCM-Based Battery Thermal Management System for Lithium-Ion Battery Modules: Experimental Investigation and Thermal Optimization
Abstract
To ensure safe operating temperatures in lithium-ion battery modules, which can enhance the performance, reliability and lifetime of electric vehicles. In this study, a hybrid battery thermal management system (BTMS) incorporating CuO (copper oxide) enhanced paraffin phase change material (PCM) and real-time thermal monitoring and predictive analytics is presented. A 4S lithium-ion battery module consisting of 18650 cells were experimentally tested by repeated charge–discharge cycles. The temperature, voltage, current and state of charge were monitored continuously using an ESP32 based IoT platform and a hybrid XGBoost–ANN model was developed for temperature and phase-state prediction. The proposed model gave the prediction accuracy of 96.7%, RMSE equal to 0.71 °C and R 2 of 0.945. Optimal thickness and loading of PCM for minimization of thermal overshoot were obtained by optimizing using GA, which was 7 mm and 25% respectively. The CuO modified PCM showed a reduction in the peak battery temperature by 25.5% and an improvement in thermal uniformity during fast charging. The developed system also exhibited stable thermal performance under long cycling conditions and provided for quick detection of thermal anomalies using IoT monitoring. The results show that the suggested hybrid BTMS is a suitable and flexible solution for high power lithium-ion battery applications.