Power transformer working state monitoring method and system

Abstract

The invention discloses a power transformer working state monitoring method and system. The method comprises the steps that acoustic signals of multiple monitoring points of a transformer are acquiredso that multiple transformer acoustic signal time sequences are obtained; sliding window segmentation is performed on the transformer acoustic signal time sequences so that multiple transformer acoustic signal time sequence sub-segments are obtained, and the correlation coefficient matrix of each transformer acoustic signal time sequence sub-segment is calculated; the singular spectrum entropy ofeach correlation coefficient matrix is calculated, and all the singular spectrum entropies are fused so that a singular spectrum entropy fusion curve of the transformer acoustic signal time sequencesis obtained; the weighted grey correlation degree of the singular spectrum entropy fusion curve of the transformer acoustic signal time sequences and the singular spectrum entropy fusion curve of thecorresponding transformer historical acoustic signal time sequences is calculated; and the working state of the transformer is determined by the weighted grey correlation degree. The monitoring efficiency and the monitoring accuracy of the working state of the power transformer can be enhanced.

Description

technical field [0001] The invention relates to the field of signal monitoring, in particular to a method and system for monitoring the working state of a power transformer. Background technique [0002] As a primary device for voltage conversion and power transmission, the power transformer plays a very important role in the power system and is called the heart of the power system. Its safe and stable operation is of great significance to the reliability and stability of the power grid. The transformers in the power system are used in large quantities, with various capacity levels and specifications. Faults and accidents during long-term operation cannot be completely avoided. Once a transformer fails, it may bring huge economic losses to the power grid and endanger operation and maintenance personnel. Personal safety. Therefore, timely and effective monitoring and analysis of the working status of power transformers in operation, early detection of potential faults and hi...

AI Technical Summary

Problems solved by technology

However, the existing mainstream analysis methods of transformer acoustic signals, such as fast Fourier transform and wavelet analysis, mostly analyze the acoustic signal of a single monitoring point, and the efficiency is low; the fast Fourier algorithm is difficult to analyze the non-stationary fault acoustic signal, wavelet The analysis will be limited by the selected wavelet base, and the accuracy is low

Therefore, the existing transformer monitoring methods cannot realize efficient and accurate judgment of the working state of the transformer

Images