Acoustic detection technology for gas pipeline leakage.

Gas leakage from pipeline leads to significant environmental damages and industrial hazards, so small leakage detection for gas pipeline is essential to avoid these serious leakages. However, because of the high frequency component of leakage signal attenuates quickly, traditional detection method which inspects pressure or vibration signal has problem to get effective information from leakage signal. So, a novel detection method based on acoustic wave is proposed. This paper, firstly, researches on the phonation principle of pipeline leakage and the characteristic of sound source, and simulates the leakage acoustic field on the basis of aero acoustics. Secondly, using Wavelet Packet Transform method and Fuzzy Support Vector Machine pattern classification, the laboratory testing for identifying acoustic signal of gas pipeline leakage is presented. Finally, the field application demonstrates that the detection system could identify small gas leakage effectively and avoids false-alarms which caused by running conditions with a good prospect.