Pressure-Leak Duality for Leak Detection and Localization in Water Distribution Systems

Water utilities are challenged to reduce their water losses through detecting, localizing, and repairing leaks as quickly possible in aging distribution systems. In this work, we solve challenging problem by detecting multiple simultaneously a network for the Battle of Leak Detection Isolation Methods. The performance leak detection localization depends on how well system roughness demand calibrated. addition, existing affect diagnosis unless they identified explicitly represented model. To circumvent chicken-and-egg dilemma, decompose into levels decision-making (a hierarchical approach) where iteratively improve model so able multileak problem. First, combination time series cluster analysis is used smart meter data build patterns models. Second, point interval estimates pipe roughnesses retrieved using least squares calibrate hydraulic model, utilizing models from first step. Finally, calibrated primal transformed dual that intrinsically combines sensor hydraulics. This automatically converts small pressure deviations caused sharp localized signals form virtual flows. Analytical derivations sensitivities with respect these flows calculated estimate leakage impulse responses at candidate nodes. Subsequently, use (1) detect start leaks, (2) compute Pearson correlation residuals, which allows further leaks. novel modeling approach resulted highest true-positive rates isolation among all participating teams competition.