Optimization of Water Supply Network Design Based on a Tabu Search Algorithm

The optimal design of a water supply system is essential to minimize costs as well as to maximize the sustainability of the infrastructure. Moreover, most future problems related to its operation and possible extensions can be significantly reduced at the design stage. The network performance is subject to laws of flow balance and energy conservation, which represent non-linear constraints. Furthermore, the pipe diameters, which have to be chosen from a commercial catalog, are discrete variables. Consequently, it is a non-linear mixed-integer problem whose resolution is not trivial. For this type of problem, metaheuristics are suitable resolution methods because of their abilities to explore the search space in order to find an optimal solution. In this study, a tabu search algorithm is used to design the well-known Alperovits and Shamir’s network. Although the global optimum is not reached, the efficacy of the proposed method is fully demonstrated. The results are similar to those achieved by other metaheuristics and the resolution times are considerably shorter. Therefore, it is expected to obtain better results in future studies in which larger networks are designed.