Central Force Optimization Method Used to Optimize Water Distribution Networks

Water supply networks account for a major portion of the urban infrastructure requiring sizable funds for their construction and operation. This has encouraged many researchers to focus their efforts on optimizing water distribution networks. In most recent research in the field, stochastic meta-heuristic algorithms have been employed to address the problem. Given the many inadequacies of stochastic algorithms, the present study explored the application of a novel deterministic algorithm for minimizing pipe-sizing costs in water distribution systems (WDS). The algorithm used in this study is a modified version of the central force optimization algorithm called CFONet which is used for solving water distribution system problems. The performance of the proposed method was evaluated by optimizing pipe-sizing in two benchmark WDSs. For this purpose, both CFO and CFONet methods were written in MATLAB and interfaced with the hydraulic simulation model EPANET. Application of the two methods mentioned above to two different water supply systems revealed the superior performance of the proposed CFOnet over its CFO counterpart. Thus, CFOnet yielded an optimum response for the deisgn of a double ringed system equal to US$419,000 after 12,432 runs of objective function estimations. Moreover, the optimum design of a CAdo system was obtained to cost 126,535,915 In. Ruppes after 259,476 runs of objective function estimations. Comparison of the results obtained in this study with those reported in the literature on other stochastic optimization algorithms showed that the proposed method yields satisfactory solutions for WDSs optimization problems while it also enjoys the merits of a deterministic optimization method.