Optimum Design of Water Networks Using A Micro Software Based Genetic Algorithms Multiobjective

Water distribution networks are a system that searches for water and brings it to each user according to what they use the water for. The infrastructure needed to achieve this goal requires significant financial investment; therefore, providing the optimal design of these networks is very important. Not only is the economic aspect essential; in addition to it, there are others such as hydraulic behaviour, which refers to maximum pressures and speeds, availability of pipes, water quality, demand distribution, network reliability, system performance which further complicates an integral analysis leading to an optimal design. The optimization of a water network, due to its complexity, is associated to the combinatorial problem called NP-HARD, which means that it is not possible to use a deterministic method to solve it, but requires special methodologies that make use of a reasonable computational processing time. This is necessary to obtain a configuration of diameters complying with the limitations (speed and pressure), as well as achieving acceptable values in the objectives (lower cost and higher reliability). Using multiobjective genetic algorithms can solve this type of NP-HARD combinatorial problems; and consequently, optimizes the water network. To find an optimal set of solutions, we developed a computer program called Magmoredes, which based on the use of a multiobjective genetic microarray, proposed by Coello and Toscano [1], adapted for the application of water distribution networks. To comply the economic demands, under pressure and speed limits according to Peruvian regulations a program in the Java programming language was developed. Finally, verification of the proposed algorithm on the efficiency of the Hanoi water network is performed (Fujiwara and Khang, [2]). This water network has a single source with 3 basic circuits, 31 nodes, 1 reservoir and 34 pipes. The nodes are at the same elevation and minor losses in the pipes are not considered.