Low-Voltage Series Arc Fault Detection Method Based on VMD

Aiming at the current situation that the traditional overcurrent protection device is difficult to detect the low-voltage series arc fault, a detection model for low-voltage series fault arc based on Variational Mode Decomposition (VMD) was proposed, utilizing a VMD method optimized by the Whale Optimization Algorithm (WOA). Firstly, WOA was used to optimize the VMD parameters of different loads to decompose the signal. Then, a multi-feature data set was constructed by combining Pearson correlation coefficient and random forest importance assessment. Finally, combined with support vector machine, a low-voltage series arc fault detection model was built. The results show that the accuracy of the proposed arc fault detection method is more than 99 %, and it is suitable for a variety of typical loads.