Acoustic emission source identification technique for buried gas pipeline leak

Leaks in gas pipelines cause unnecessary waste of limited resources and produce danger factors, thus leak testing is necessary. Acoustic emission (AE) technology is one of the promising methods for pipeline leak testing. AE signals of pipeline leak carry the feature information of structure integrity (the dimension and location of leak source, etc.), which are stochastic and uncertain, and belongs to non-stationary signals. Because of the noise and the complexity of AE signal transmission, the identification of AE source is very difficult. On the basis of analyzing the characteristics of AE signal and background noise, we established a gas leak identification model for city gas pipeline in this paper. A leak identification method is presented based on spatial-temporal data fusion. The multi-data segments are fused in time and space will decrease incertitude in the process of identification. Experimental result shows that the inspection range can be up to 87m, and the identification rate can be up to 95% for phi 1mm pinhole leak.