DC Arc Fault Detector Based on Wavelet Transform for DC Microgrid

In recent years, with the increasing capacity of power generation from renewable energy in the power grid, the development of a DC microgrid system is dramatical. In a DC microgrid system, an arc fault can be caused by loosening connection of conductors, aging and damage of insulations. It is an extremely harmful potential fault in a DC microgrid. But, a reliable arc fault detector for the DC microgrid is still lacking. A DC microgrid composed of a large number of branches has a complex structure and the operation condition varies with the branches. Though there are existing DC arc-fault detectors, most of them are developed for DC systems with a simple structure and limited operating conditions. There are few mature industrial products for the DC microgrid. Therefore, it is significant to study the characteristics of DC arc faults and develop a reliable arc fault detector for DC microgrids. In this paper, the potential arc faults in a DC microgrid are emulated. The characteristics of DC arc current in the time domain, frequency domain, and time-frequency domain are investigated. According to the characteristics of arc faults, the wavelet transform is evaluated to extract the arc fault features in the time-frequency domain. In this way, a cost-effective arc fault detector is developed for DC microgrids. It can improve the safety and efficiency of a DC microgrid.