Forming District Metered Areas in a Water Distribution Network Using Genetic Algorithms

Abstract

Dividing a water distribution network into District Metered Areas (DMAs) is the first vital step towards water loss management. It is a multivariate problem. Techniques based on Genetic Algorithms is a proven way to provide a very good solution to optimization problems. Basic requirement is that each problem must be well described by an objective function. The formation of the objective function is tested through Matlab's optimization tools. The results of the genetic algorithm are compared to the results of an algorithm (in C++ Language) developed in an earlier stage to provide optimal system pressure reduction by closing indicated pipes. The process for the formation of the objective function and a case study on a real network are presented.