A Particle Swarm Optimization (PSO) heuristic for water distribution system analysis

Looped water distribution systems are conventionally analyzed by using iterative numerical methods. The aim of the applied numerical methods is to solve the system of non-linear equations which express the energy and mass balance at the nodes. Instead of the widely used numerical solution of these non-linear equations, this paper introduces two heuristic optimization-based models for the hydraulic analysis of the water distribution networks. The optimization is achieved by using a modified Particle Swarm Optimization method which is a Heuristic method. The analysis is based on the h-Equation and the Darcy-Weisbach head loss equation. The first method modulates a cocontent objective function based on the energy fluxes. The second approach, which is presented for the first time in this article, is based on the objective function which directly checks the simultaneous validity of the mass balance equations at the nodes. Both cases end to an unconstrained optimization problem. Since the Darcy-Weisbach head loss is selected, the hydraulic resistances do not remain constant at each iteration, and therefore, a non-conventional optimization method must be applied. By comparing the two presented methods, the one based on the check of the simultaneous validity of the mass balance equations at the nodes better. This method could also be used for sensitivity analysis and for a more realistic investigation of the water distribution networks.