A Novel Leak Localization Method for Water Pipeline Systems Based on Acoustic Emission Monitoring and Event Correlation

This paper proposes a novel method for leak localization in water pipeline systems through the monitoring of acoustic emissions and cross-correlation of acoustic emission events. The method is designed to overcome limitations in the calculation of the time difference of arrival by traditional methods such as time-of-arrival detection and cross-correlation. The proposed method combines these two conventional techniques, specifically by examining cross-correlation at positions associated with arrived acoustic emission events. The positions of acoustic emission events are initially identified using the constant false alarm rate signal detection algorithm. Subsequently, the minimum entropy deconvolution is applied at these positions to eliminate noise and reveal signal information. Finally, cross-correlation at the identified positions is computed to determine the time difference of arrival and localize the leak position. Experimental validations on a real-world testbed demonstrate the effectiveness of the proposed method with a high level of accuracy. The proposed method exhibits superior accuracy in comparison to other methods.