Automatic Creation of District Metered Areas in Urban Water Distribution Networks Using Community Structure Algorithm and Genetic Algorithm aiming at Equitable Distribution of Network Pressure

A District Metered Area (DMA) is a specified area in which inlet and outlet water at any time is measured. The procedure of forming District Metered Areas is generally comprised of two phases: The clustering phase for automatic detection of the best communities and the physical partitioning phase to optimize the location of gate valves and flowmeters according to the objective function. In this study, the network of Taft city in Yazd province was modeled in Epanet software and by linking EPANET and MATLAB and loading the network configurations, the network was automatically clustered using community structure. In the physical partitioning phase, using the genetic algorithm, the optimal locations of the gate valves and flowmeters was determined by considering the objective function of reduction of standard deviation of DMAs average pressure. The results showed that the community structure algorithm with average daily pressure weight can automatically create appropriate clustering according to the modularity index and uniform pressure in clusters. The optimization of boundary pipes resulted in the reduction of the pressure in different parts of the network, such as areas with pressure higher than the average pressure of the network, and due to the reduced standard deviation of DMAs average pressure caused equitable distribution of network pressure.