Leak detection in drinking water network using pressure-based classifier

Water leakage in drinking water networks (DWN) is the main reason for water loss in water networks. Considering the worldwide problem of water scarcity added to the challenges that a growing population brings, minimizing the water losses through detection of water leakages in DWN using efficient techniques is an urgent humanitarian need. This paper proposes a new data-driven leak detection method of defining classifiers based on limit pressure measurements in DWNs. This can get rid of complexities and application constraints of the model-based approach. The end result is an average 60% detection accuracy with limited data requirements. The proposed approach is applied to Hanoi DWN.Peer ReviewedPostprint (author's final draft