Traveling wave fault location based on wavelet and improved singular value difference spectrum

Abstract

The single terminal fault location method is based on the wavelet modulus maximum to obtain the time difference that the fault traveling wave reach the same measuring point, which can calculate the fault distance, but the error of the actual results is relatively large. In this paper, a traveling wave fault location method based on Wavelet and improved singular value difference spectrum is proposed. Firstly, the delay step of the time series is determined by the autocorrelation function of the fault traveling wave signal, and construct the new Hankel matrix. Secondly, the singular value decomposition method is used to get the new matrix singular value number, and the number of effective singular values is calculated by the singular value difference spectral, the singular values are reconstructed to get the de-noising signals. At last, using the modulus maximum to get the changed frequency, the fault distance can be calculated according to the relationship between the frequency and the fault distance. The simulation experiments demonstrate that the error of the proposed method is small and the accuracy is higher.