Developed Swarm Optimizer: A New Method for Sizing Optimization of Water Distribution Systems

Abstract

The introduction of metaheuristic algorithms in water resources engineering has greatly raised the need for continued development of appropriate optimization methodologies for analysis, planning, design, and operation of water resources systems. This paper proposes a novel developed swarm-based optimization algorithm named DSO, which integrates the accelerated particle swarm optimization (PSO) with the big bang-big crunch algorithm (BB-BC) to optimize the design of water distribution systems (WDSs). Traditional PSO is easy to fall into stagnation when no particle explores a position that is better than its previous best position for several iterations. To deal with the problem of maintaining diversity within the swarm and to enhance the exploration in the search, the concepts of the Big Crunch and Big Bang strategies from the BB-BC algorithm are incorporated into the global and local searching steps of the accelerated PSO, respectively. In addition, a harmony search–based strategy is used to control the location of generated particles, and finally a modified version of the feasible-based mechanism is applied to handle the constraints. The DSO approach obtains competitive results on three well-known benchmark WDS optimization problems, with a number of decision variables ranging from 30 to 454, at a relatively low computational cost.