The Application of An Improved Particle Swarm Optimization Algorithm in Design of Water Distribution Systems

Optimal design of water distribution system is a real problem and its aim is finding the best solution for transferring of water from the reservoir to consumers;so that, all their needs to be supplied with the lowest cost. Many researchers proposed various researches such as linear, nonlinear programming, global optimization,and evolutionary algorithms for minimizing the cost of these systems. The Particle Swarm Optimization (PSO) algorithm, which is an evolutionary algorithm originally introduced for solving optimization problem with continuous variables, but later it was developed for problems with discretevariables. Inthis research, for minimizing the cost of waterdistribution system design after modifications in continuousPSO algorithm, the program was linked to the hydraulic simulator (EPANET 2.0 in Matlab environment). Finally, for improving the algorithm performance, and obtaining the best values of PSO in optimization and water distribution system design, we were used three benchmark networks from previous studies that are different in layout and also the number of pipes, which they called Two-loop, Hanoi, and Kadu networks. From Two-loop network optimization, the least-cost network was estimated from 3100 number of objective function evaluation, which caused to improve the results, when compared to the pervious algorithms. In Hanoi network, the minimum cost was obtained $6.097×106 with =10.667 value, which was less than the results of other networks. Additionally, in Kadu network the cost was obtained 130,666,043 Indian rupees, which was less when compare to the other methods. The results showed a good performance of this algorithm when compared to the other algorithms.