Development and implementation of a leak detection model with the genetic algorithm

ABSTRACT The aim of this study is to propose a two-step solution methodology based on hydraulic modeling and optimization algorithms using pressure residuals for leak detection water distribution networks (WDNs). In the first step, a leakage detection model is created by performing a series of hydraulic simulations. In the second step, the leak detection problem is defined in accordance with an leakage detection model (LDM)-based solution and a method that will perform the leak detection in WDNs faster and more precisely. The model was tested on a hypothetical network designed for this study and the well-known Hanoi network, a widely used benchmark in hydraulic studies, frequently employed for performance comparison in leak detection research. The cumulative correct result rate is 92%. In the hypothetical network, the most marked node matched with the real leakage location after running the algorithm for a maximum of five times. The method may reduce economic losses by shortening detection times.