Optimal Design of Water Distribution Network under Hydraulic Uncertainties and Robusting

Uncertainty in the planned water demand and changes in the roughness values ​​of the pipes are among the common cases of unfavorable performance of water distribution networks. In this research, a model has been developed to optimize networks that have been implemented but need to be strengthened under conditions of uncertainty. In this model, the EPANET simulator model was combined with optimization genetic algorithm in MATLAB programming environment. In such a way that it is first created using the concept of fuzzy logic, fuzzy membership functions of the input parameters(nodal demand and pipe roughness coefficient), then by considering the relationship between demands and pressure head requirements at nodes and using (Gupta and Bhave 2007) impact method The fuzzy optimization model is converted to a deterministic model and finally, the diameter of the pipes that should be installed parallel to the existing pipes is obtained by using the genetic algorithm. The method created for a typical network from past researches is checked by changing the uncertainty level, the results indicated a 28% increase in the cost of the network with a 50% increase in considered uncertainties and a 7.5% decrease in cost by applying a tolerance of 0.75 meters in the desirable pressure head. The proposed method based on the genetic algorithm is completely suitable and creates a solution that provides the required pressures in the worst situation variables.