Detection of Arc Fault by using ML Algorithm

In this paper, a novel machine learning (ML) technique is introduced for the detection of arc defects in electrical circuits. Signals from various electrical loads were examined for statistical characteristics that describe signal anomalies, such as standard deviation, skewness, and kurtosis. For accurate arc defect identification, a Random Forest-based machine learning model was developed. The model showed excellent accuracy in identifying healthy and fault electrical signals. Its performance was verified using real-world datasets, demonstrating its application in real-time electrical systems. This study makes a positive impact on electrical system safety by quickly and precisely identifying arc defects using ML algorithms. It emphasizes how ML-based methods might advance the diagnosis and prevention of electrical faults.