Research On Leakage Detection Via Hydraulic Model Calibration In Water Distribution Systems

Humans are faced with serious water shortage problems increasingly. reducing theloss of water from distribution systems not only can save water resources, at the sametime can improve the water supply enterprise’s service level. Water suppliers areincreasingly under pressure to continue to reduce the loss of water from distributionsystems. Therefore, How to reduce the water leakage is a problem need to be solvedurgently.Water supply network of our country is complex and the leakage point positioningis difficult. the leakage point location problem plagued water supply industry for a longtime. This paper presents a modelbased optimization method for leakage detection ofwater distribution systems. Leakage hotspots are assumed to exist at the model nodesidentified. Leakage is represented as pressure-dependent demand simulated as emitterflows at selected model nodes. The leakage detection method is formulated to optimizethe leakage node locations and their associated emitter coefficients such that thedifferences between the model predicted and the field observed values for pressure andflow are minimized.This paper a novel evolutionary algorithm, suitable for continuous nonlinearoptimization problems is introduced. This optimization algorithm is inspired by the lifeof a bird family, called Cuckoo. The algorithm has the following features:(1) Has fastconvergence speed, the termination of the number of iterations significantly less thangenetic algorithm and particle swarm algorithm, calculation efficiency is four timesbigger than the genetic algorithm and particle swarm algorithm.(2) Calculation processcan be timely release the computer memory, memory is less. As a result, the cuckoooptimization algorithm is superior to genetic algorithm and particle swarm algorithm.Application of the proposed algorithm to some benchmark functions and two realengineering examples has proven its capability to deal with difficult optimizationproblems.This paper adopts two engineering examples to illustrate the leakage detectionmethod in actual pipe network.(1) To determine the leakage areathrough the NightMinimum Flow analysis.(2) Comparison and analysis the change of the meter, find themeter which flow rate changes significantly, drawing draw the water meter water path,the path has new leakages, narrowed the leakage area.(3) Computing node emittercoefficient through pressure-dependent leakage detection model, leakage hot position isdetermined, further narrowing the leakage area.(4) Detecting the pipeline near thehotspots by using of leakage detecting instrument to determine the accurate position ofthe leakage. Engineering examples have proven that the pressure-dependment leak detectionapproach is effective at identifying likely leakage hotspots, provided the field data isavailable at MNF hours.The pressure-dependment leak detection approach is2timesmore efficient than sweep-through by sounding alone.