A Study on Characteristics of Arc Fault of Electrical Load Based on RS485 Communication

An arc fault because of a damaged electrical wire may cause an electrical fire. The ability of arc faults to cause electrical fires has necessitated research and development in the field of arc fault detection. This thesis studies and analyzes arc faults at a low voltage level in the presence of household electrical loads. The thesis will provide a component-based approach to characterizing domestic electrical loads and to observe and process the signals and characteristics of these loads in the presence and absence of a series arc fault. Arc waveform characteristics can be evaluated with various methods to recognize the presence of hazardous arc fault conditions. This thesis explores a method of processing the low-frequency signals of electric currents, a method of detecting the arc signals using FM IF system’s RSSI from certain frequency band signals in high-frequency noise signals arising when arcs occur, and an algorithm for processing the arc signals. The results will be used to conclude the behavior of different load types and whether loads falling under the same category show similar characteristics. These results are intended to assist in improving the detection of series arc faults in the presence of household electrical loads. The data obtained was analyzed using signal STFT and CWT. For all tested load categories, the characteristic features of the arc were analyzed. The CWT had better characterization capability due to its resolution.