Acoustic emission detection method for gas pipeline leakage based on AI chip

Abstract

The invention relates to a gas pipeline leakage acoustic emission detection method based on an AI chip, and the method comprises the following steps: obtaining sound signals collected by N sound transduction devices, N > = 2, and extracting feature vectors from the sound signals; inputting the extracted feature vectors into a convolutional neural network to obtain qualitative classification of pipeline leakage; the leakage indication number is counted according to qualitative classification of pipeline leakage, and when the leakage indication number exceeds a preset threshold value, the sound transduction device transmits early warning data and sound waveform data corresponding to leakage; and when any two adjacent sound transduction devices transmit early warning data, performing empirical mode decomposition on sound waveform data transmitted by the two sound transduction devices, and calculating the position of the pipeline leakage point. According to the method, the sound signals monitored by the sound energy conversion devices are processed in a mode of implementing the deep learning model, so that the position of the pipeline leakage point is determined through the sound signals of the two adjacent sound energy conversion devices.

Description

Technical field [0001] The present invention involves the field of safety technology in the gas pipe network, and specially involves a sound launch detection method based on AI chip gas pipeline leakage. Background technique [0002] Gas pipeline leak detection is an important part of ensuring gas safety production. At present, common methods include stress -based methods and sound -based methods. The pressure -based method is represented by a negative pressure wave. The negative pressure wave refers to a negative pressure fluctuation generated by the pipeline leak. The leakage will reduce the pressure of the leakage point. The pressure sensor detects the transient pressure drop shape to determine whether the leakage is leaked, and the position of the leak is determined. The sound wave -based method is represented by sound emission detection. When the pipeline leaks, the fluid is sprayed outward through cracks or pores to form sound sources. The monitoring of sound waves is to co...

AI Technical Summary

Problems solved by technology

This method is only suitable for some scenarios, because the gas pipeline environment is complex and changeable, and useful signals and unwanted signals may be in different frequency bands or in the same frequency band, so it is difficult to directly judge signals

Images