Switching Characteristics of Domestic Loads for Reliable Arc Fault Detection

The increasing complexity of modern domestic loads, driven by advances in power electronics and control mechanisms, poses significant challenges for arc fault detection. This paper presents a comprehensive analysis of switching characteristics in domestic appliances during both OFF-to-ON and ON-to-OFF transitions. A real-time experimental setup was developed to collect current waveforms in 16 representative loads under controlled laboratory conditions. Transient signals were analyzed using four statistical features: mean, standard deviation, skewness, and kurtosis. Based on these descriptors, a K-Means clustering algorithm was used to group the loads according to their switching behavior. The results provide critical insights into the statistical separability of load types and contribute to the development of more robust and data-driven arc fault detection frameworks for low-voltage residential systems.