In-depth learning-based partial discharge ultrasonic audio frequency identification method and system

Abstract

The invention provides an in-depth learning-based partial discharge ultrasonic audio frequency identification method and system. The method comprises the following steps: partial discharge ultrasonic signals of an electric power device are detected, partial discharge ultrasonic audio frequency data is obtained and then converted into a sound spectrogram, a depth convolution nerve network model is built, samples are used for training a network, partial discharge ultrasonic audio frequency data to be diagnosed is input into a trained network, and a partial discharge defect type is output and obtained. According to the in-depth learning-based partial discharge ultrasonic audio frequency identification method and system, the partial discharge ultrasonic audio frequency data is converted into the sound spectrogram, a depth convolution nerve network is used for identifying the sound spectrogram, ultrasonic signals of all kinds of defects in partial discharge can be accurately and effectively identified, and a convenient and reliable diagnosis method is provided for electric power device insulation state assessment.

Description

technical field [0001] The invention relates to the field of fault diagnosis of power equipment, in particular, to a partial discharge ultrasonic audio recognition method and system based on deep learning. Background technique [0002] Partial discharge will generate acoustic signals inside the power equipment, and the ultrasonic method measures the partial discharge signal by installing an ultrasonic sensor on the outer wall of the equipment cavity. The method is characterized by no electrical interference and high positioning accuracy. By collecting, analyzing and judging acoustic signals, recording ultrasonic data of various partial discharges, analyzing the characteristic differences of ultrasonic signals of various partial discharges, and judging the types of partial discharges, it is possible to identify and locate partial discharge faults of power equipment, and to maintain power equipment Safe and stable operation is guaranteed. [0003] Currently commonly used ult...

AI Technical Summary

Problems solved by technology

[0003] Currently commonly used ultrasonic diagnostic methods have limitations. For example, the judgment rules of the ultrasonic amplitude threshold judgment method are simple and cannot identify defect types. For example, the phase map method has limited characteristic parameters and low accuracy. For example, the traditional speech recognition method has high training complexity and difficult decoding. High latency

Images