Abstract:

However,the existing DC arc fault detection methods are easy to be affected by the environment,which can not achieve accurate detection,and the core algorithm is complex,difficult to deploy in practical applications. To solve these problems,a DC arc fault detection algorithm based on short-time Fourier transform is designed. By analyzing the amplitude variation trend of arc current signal in frequency domain,the spectrum is divided into different stages,the ratio of different stages is calculated,and the threshold value is obtained according to a large number of experimental results to realize arc fault detection. Finally,a suitable single chip microcomputer is selected to implement the detection algorithm and tested to verify the effectiveness of the proposed algorithm. The results show that the algorithm is not only able to detect arcing faults in a timely manner with multiple types of loads connected in parallel,the accuracy of detection can reach 94.4%. In addition,the algorithm has a misjudgment rate of 0 in the face of normal state currents,which indicates that it has good robustness.

Key words: DC arc fault, short-time Fourier transform, time-frequency domain data analysis, arc fault detection