Leak Identification Method for Buried Gas Pipeline Based on Spatial-Temporal Data Fusion

Acoustic emission (AE) signal of pipeline leak carries the feature information of structure integrity (the dimension and location of leak source), whereas it is stochastic and uncertain. It belongs to random and non-stationary signals in its statistics. Because of the background noise and the complexity of AE signal transmission, it is very difficult to identify the AE source. Presented in this paper is a method to identify a pipeline leak based on spatial-temporal data fusion model. To elucidate the role of this method, the how-to of completion of hardware and software processing is explained in detail. Fusion of multi-data segments in time and space will decrease uncertainty in the process of identification. Experimental results show that the length of testing pipeline can be up to 87 m, and the identification rate can be up to 95% for Phi 1 mm and above Phi 1 mm pinhole leak.