NCR-Enabled Digital Twin Based EV Fire Prevention System Development

초록

In this study, we propose a digital twin (DT) based monitoring system that uses the network-controlled repeater (NCR) to transfer the temperature status of vehicles through wireless links with dual spectrum CCTVs. We developed an early detection method for electric vehicle (EV) fires by combining the adaptive Kalman filter (AKF), long short-term memory (LSTM), and one-class support vector machine (OCSVM). The system monitors temperature anomalies in EVs, focusing on detecting abnormal high-temperature increases that precede fire events. By utilizing dual-spectrum CCTV cameras and an AI-based model, the system predicts abnormal EV behaviors. Experimental results demonstrate that a significant reduction in false fire alarms and improved true fire alarm detection is achievable, with the overall AUC increase of approximately 6.5% when AKF was fully applied. © 2025 IEEE.

키워드