Gas pipeline leak recognition method based on convolutional neural network

Abstract

The present invention proposes a gas pipeline leakage identification method based on a convolutional neural network. After collecting the leakage acoustic signal and the background acoustic signal of typical leakage types, performing frame division processing and short-time Fourier transformation to obtain the characteristic of the original leakage acoustic signal Time-frequency diagram; then build a convolutional neural network classification model for leakage acoustic signals, and change the traditional square convolution kernel into a specific strip-shaped rectangular convolution kernel so that the lines in the time-frequency diagram can be better extracted Spectral features; the time-frequency diagram of leakage sound and background sound is mixed and sent to the built convolutional neural network for training. The training uses K-fold cross-validation to optimize the hyperparameters of the network model, so as to select the optimal model hyperparameters And enhance the robustness and universality of the model. Compared with the pipeline leakage identification method in the prior art, the method not only further improves the identification rate, but also can effectively solve the most difficult feature screening problem in the prior art.

Description

technical field [0001] The invention belongs to the field of gas pipeline leakage identification, and in particular relates to a gas pipeline leakage identification method based on a convolutional neural network. Background technique [0002] With the development of the economy and the continuous improvement of people's living standards, natural gas has been popularized in the daily life of urban and rural residents, and a large number of gas pipelines are distributed in the underground space of modern cities. With the passage of time and the development of the city, the pipelines distributed in the underground space will gradually age, corrode or be artificially damaged, which will inevitably lead to gas leakage. Gas leakage not only causes great pollution to the environment, but also poses a great hidden danger to the personal safety of urban and rural residents. Therefore, it is particularly important to detect the leakage source in time to ensure the personal safety of ...

AI Technical Summary

Problems solved by technology

[0004] In order to solve the time-consuming and labor-intensive problems of the traditional leakage identification method and the high rate of false alarms and false alarms, the present invention proposes a leakage identification method based on convolutional neural network, especially on the convolution kernel of convolutional neural network Made creative improvements to the acoustic characteristics of gas leaks

Compared with the pipeline leakage identification method in the prior art, this method not only has a further improvement in the recognition rate, but also can effectively solve the most difficult feature screening problem in the prior art.

Images