Data filtering method applied to ultrasonic partial discharge detection

Abstract

The invention discloses a data filtering method applied to ultrasonic partial discharge detection. A median filter which is an atypical nonlinear filter for completing signal recovery based on sequence statistics is used, aiming at pulse interference possibly existing in the ultrasonic partial discharge detection. Under the one-dimensional condition, the essence is a sliding window with a certainlength, and a signal at the central position of the window is replaced by a median after all values in the window are sorted. When the pulse width is less than half of the window width, the median filtering is very effective to narrow pulse interference. Then, for processing of random noise, after wavelet packet threshold function denoising and singular value decomposition algorithms are compared,a singular value decomposition filtering algorithm based on a singular value energy difference spectrum is provided. The singular value energy differential spectrum defined in the algorithm can welldetermine the order of a reconstruction matrix, so that the singular value contributed by random noise and the singular value contributed by partial discharge signals are distinguished, and a good filtering effect is achieved.

Description

technical field [0001] The invention relates to the field of digital signal filtering, in particular to a data filtering method based on ultrasonic partial discharge signals. Background technique [0002] In recent years, the development of ultrasonic detection technology has become more and more mature, and has been widely used in partial discharge signal detection, especially for partial discharge detection of transformers, GIS and high-voltage switch cabinets, which has achieved good results. Ultrasonic partial discharge signal detection first needs to collect the ultrasonic signal on site, then filter the ultrasonic signal, then analyze the signal, and finally judge the type of partial discharge and even locate the source of partial discharge. The judgment of the PD type and the localization of the PD source have a great relationship with the filtering of the ultrasonic PD signal. A good filter is conducive to the subsequent work, on the contrary, if there is no good fi...

AI Technical Summary

Problems solved by technology

However, due to the complexity of the ultrasonic signal, although the noise reduction method based on singular value decomposition has a good effect on random noise, how to choose the order of the reconstruction matrix is ​​a difficult point. The selection of the matrix order directly affects the filter Effect

Images