Water distribution networks have been always very important as infrastructures. The duty of water distribution networks is to supply treated water to each of the users with enough quantity, good quality and standard pressure. The quality and efficiency of these networks effect on the people’s life conditions, health and sanitary and also on continuance of economic affairs. The desires that cause the engineers to design economic networks. Optimal design is considered very important from few past decades until now. The first priority that is considered in optimum designing of the networks. Nowadays, the other important goal in designing of water distribution networks is their reliability. Economic design of water distribution networks is that it costs at least as a first priority in the design of the network considered. Today, an important objective in the design of water supply networks in addition minimizes the high costs, should be considered, and is the reliability of the network. The aim of this research is two-objective optimization of water networks by introducing the reliability as a new index by using genetic algorithm (NSGA-II) for optimization and EPANET as a simulator. In this study, a new criterion is introduced to iect the mechanical y and hydraulic reliability at the same time. For the hydraulic reliability, Prasad and Park index (In) is considered and for mechanical reliability, entropy index (e) is considered. Objective function is to minimize the costs and maximize the network reliability. These two aims are included of the answers that one of them is selected based of the budget or the expert’s opinion. The violation functions have been used for pressure and velocity. In this study, the violation functions are described based on justify; LINE-HEIGHT: normal; TEXT-INDENT: 0.25in; MARGIN: 0in 0in 0pt" Keywords : Reliability, Mechanical reliability, Hydraulic reliability, Entropy, EPANET, Water distribution networks, Multi-objective optimization, Fuzzy logic, Genetic algorithms
