The detection of small leak of transportation pipes is significantly important for pipe safety. Most techniques have solved leak detection through installing sensors, in which small leak is still a challenge because of its tiny change. To address small leak detection of gas transportation pipes, Gaussian-based models are proposed to learn the distribution of small leak acoustic signals. For transportation pipes, acoustic signals of small leak are combined with environmentally and randomly high noise, which increases the difficulty in learning the acoustic features, especially based on limited data sets. After analyzing the acoustic signals, we find out that the noises and small leak signals are following certain Gaussian distribution. Therefore, in the proposed model, we establish Gaussian models using built distributions of small leak with different location and gas positions. Additionally, an acoustic signal pre-processing scheme is designed to deal with original collected signals based on power spectrum analysis. Experimental results show the proposed models perform satisfiedly with limited data. We further analyze the inherent properties of small leak of transportation pipe in simulation, and discuss the influence from leak position and gas pressure of transportation pipes.
[1]
Amir Mostafapour,et al.
Gas leak locating in steel pipe using wavelet transform and cross-correlation method
,
2014
.
[2]
Qiong Ai,et al.
Acoustic emission of fatigue crack in pressure pipe under cyclic pressure
,
2010
.
[3]
Amir Mostafapour,et al.
Leakage Locating in Underground High Pressure Gas Pipe by Acoustic Emission Method
,
2013
.
[4]
Fei Wang,et al.
Natural gas pipeline small leakage feature extraction and recognition based on LMD envelope spectrum entropy and SVM
,
2014
.
[5]
Michael Corsar,et al.
Monitoring the presence of water and water–sand droplets in a horizontal pipe with Acoustic Emission technology
,
2014
.
[6]
T. Yan,et al.
Identification of minimum transport condition for sand in two-phase flow using acoustic emission technology
,
2013
.