@article{article_1684636, title={Enhancing Reliability of EV Charging Systems: Open-Circuit Fault Detection Using Advanced Signal Processing and Machine Learning}, journal={International Journal of Engineering Research and Development}, volume={17}, pages={525–536}, year={2025}, DOI={10.29137/ijerad.1684636}, author={Okumuş, Hatice and Mollahasanoğlu, Merve and Ergün, Ebru}, keywords={Electric Vehicle, Power Electronics Fault Detection, Machine Learning, k-Nearest Neighbor, Random Forest}, abstract={The environmentally friendly nature and low operational costs of electric vehicles (EVs) have significantly increased their adoption in recent years. As EV usage grows, the need for reliable and efficient power electronics in charging infrastructure becomes more critical. These systems rely on switching components such as IGBTs and SiC MOSFETs, which are prone to failure due to high operating temperatures and currents. To address this issue, this study proposes a novel method for detecting open-circuit faults in AC/DC rectifiers used in EV charging stations. The approach analyzes three-phase current signals on the AC side to identify faulty switching devices. Feature extraction is performed using a hybrid technique that combines Discrete Wavelet Transform (DWT) and the Teager–Kaiser Energy Operator (TKEO), capturing transient fault-related characteristics. The extracted features are then classified using k-Nearest Neighbors (k-NN) and Random Forest (RF) algorithms. Performance was evaluated using 10-fold and 5-fold cross-validation. In both settings, RF outperformed k-NN across all metrics. Under 5-fold validation, RF achieved accuracy, recall, precision, and F1-score values of 0.9933, 0.9933, 0.9935, and 0.9933, respectively. These results confirm the robustness and effectiveness of the RF-based method for fault detection, making it a promising tool for predictive maintenance and fault-tolerant EV charging systems.}, number={3}, publisher={Kirikkale University}