Transformer live detection method based on dynamic time algorithm

Abstract

The invention discloses a transformer live detection method based on a dynamic time algorithm and relates to the technical field of transformer fault diagnosis. The transformer live detection method disclosed by the invention comprises the following steps: by utilizing an acoustic vibration array, acquiring acoustic vibration signals of a transformer running under multiple normal conditions, and preprocessing the acquired acoustic vibration signals of the transformer by utilizing a spectral subtraction algorithm, thus obtaining pure acoustic vibration signals; extracting characteristic quantities, namely template characteristics, from the pure acoustic vibration signals by virtue of a Mel frequency cepstrum coefficient established on the basis of Fourier and cepstrum analysis; detecting anacoustic vibration signal of a to-be-detected transformer by adopting a signal-to-noise ratio management spectral subtraction algorithm; extracting the characteristic quantities by utilizing the samemethod; and performing similarity comparison with the template characteristics by utilizing a time sequence of characteristic vectors of the acoustic signal, and performing statistics on the characteristic quantities with the highest similarity, thus a recognition result is obtained, and the problem that speed of recognizing an acoustic signal of a transformer broken down is low is effectively solved.

Description

technical field [0001] The invention belongs to the technical field of transformer fault diagnosis, and in particular relates to a transformer live detection method based on a dynamic time algorithm. Background technique [0002] Among the various equipment in the power system, the transformer is one of the more expensive and very important equipment, and its safe operation is of great significance to ensure the safety of the power grid. Based on its own price alone, the price of imported 250MVA / 500kV transformers is basically around 4 million US dollars per 3 sets, with an average of 1.33 million US dollars per set. The same specification is about 10 million yuan / set. If a large-scale power transformer has an accident during operation, it may cause a large-scale power outage, and the maintenance period generally takes more than half a year, which not only costs a lot, but also affects a wide range of areas. [0003] Our country has an early understanding of the importan...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to provide a transformer live detection method based on dynamic time algorithm, thereby overcoming the shortcomings of traditional transformer fault diagnosis methods, such as short-circuit impedance method, frequency response method, low-voltage pulse method, dissolved gas analysis method, etc. Insufficient and insufficient, and less algorithm development of transformer live detection system based on acoustic vibration array and voiceprint imaging, which brings problems of poor accuracy and slow calculation speed to transformer fault diagnosis

Images