Optimization of Eruh water distribition networks using metaheuristic methods

Due to the increasing population in our country, the correct use of water resources and the delivery of water to consumers are becoming increasingly important. The delivery of water to consumers consists of many processes that require cost. Water distribution networks constitute a large part of this cost. In this direction, the design of the network pipes we use to deliver water to consumers also gains importance. There are many methods used to design and optimize the network in order to reduce the costs of network pipes, and optimizing pipe diameters using metaheuristic methods is one of them. In this study, after an existing drinking water network line was redesigned in the Epanet program, cost optimization was made regarding the diameters of the pipes in the network using the Epanet-Matlab toolkit. In the optimization, genetic algorithm (GA) and particle swarm optimization (PSO) algorithms, which are metaheuristic methods, were used. While designing with the Epanet program, the pipe elevations and diameters of the water distribution network were prepared in accordance with the current situation. After the network lines, which were redesigned with the help of the Epanet program in accordance with the current situation, were transferred to the Matlab program, certain pressure restrictors were introduced to the network pipes. In this direction, after the network pipes were optimized using genetic algorithm and particle swarm optimization algorithms, the new pipe diameters obtained were re-entered into the Epanet program and the system was operated successfully. Optimization results with genetic algorithm and particle swarm optimization algorithms were compared. It has been seen that the pipe costs of the network have been reduced somewhat as a result of the optimizations made regarding the drinking water network line of the Eruh district of the province of Siirt, which is discussed within the scope of this thesis. When the results obtained are examined, the amount of reduction in the cost due to the variability of the control parameters of the algorithms used in the optimization also differed. In this thesis study, it was determined that successful results were obtained when the correct control parameters were used in the cost analysis using the Epanet-Matlab toolkit. Keywords: Epanet program, Matlab, Epanet-MATLAB toolset, metaheuristic optimization algorithms, Genetic algorithm, Particle swarm optimization algorithm