Time-Frequency Distribution Characteristic and Model Simulation of Photovoltaic Series Arc Fault With Power Electronic Equipment

In the photovoltaic (PV) system, the electrical characterization of series arc faults would be inevitably interfered by power electronic equipment. Existing methods are unable to effectively separate clear arc fault features from these influence factors, forming some unwanted detection barriers. In this paper, series arc fault experiments are conducted in PV systems with different kinds of power electronic equipment first. Having analyzed PV series arc faults through the existing linear transform, arc faults are hard to be identified in the undesired switching frequency band and harmonics because of noise interferences. Through establishment of effectiveness evaluation index, more performance limitations are discovered in low- and high-frequency bands for the bandpass filter effect of the linear transform. Then the arc fault identification in high and switching frequency bands become much easier after the investigation of the bilinear transform with higher time-frequency resolution. However, the performance is not obviously improved in the low-frequency band because of the cross term. Next, arc fault features in low and switching frequency bands would be more obvious through application of the adaptive transform. Besides, performances just have a slight decrease in the high-frequency band. These better time-frequency distributions are of great benefit to improve the detection accuracy and expand the application scope. Finally, the arc fault model is proposed to acquire the simulated data in the PV system with the power electronic equipment. It is found that the power electronic equipment would seriously interfere arc fault behaviors in the low and switching frequency bands.