A Robust Online Series Arc Fault Detection for Single-Phase Systems

This research focuses on real-time detection of series arc faults (SAFs) using line current analysis and load category recognition to increase reliability. These faults can be caused by loose connections or material deterioration, resulting in overheating, fire hazards, and device failure. Detecting SAF in electrical networks is difficult due to heterogeneous load categories and nonlinear currents in electronic loads, which can cause false alarms. A threshold in total harmonic distortion (THD) based on the root mean square (rms) value of the current (<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\text {THD}_{I\_{\text {RMS}}}$ </tex-math></inline-formula>) differentiates SAF occurrence in electrical and individual electronic loads. The sum of magnitudes of selected harmonic components determines the SAF in solo or combined electrical loads. The rate of change of current (<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\Delta i/{\Delta n}$ </tex-math></inline-formula>) guarantees the occurrence of SAF in combined electronic loads. The sum of selected harmonics and <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${\Delta i}/{\Delta n}$ </tex-math></inline-formula> confirm the occurrence of SAF in mixed loads. The complete experimental setup, including the proposed algorithm in NI ELVIS II, is kept under ambient atmospheric conditions. Validation yielded a high success rate of 98.8%, with reproducibility confirmed using a dataset of 14 loads and 54 combinations of normal and series arcing circumstances. The missed detection rate is 1.2%, ensuring accurate SAF detection with low errors. Usefully, the algorithm does not produce any false alarms.