عنوان مقاله [English]
The optimum design is considered as very important topic in engineering issues. Water distribution networks are one of the crucial urban infrastructures. In urban water supply systems, tanks or service reservoirs are constructed for purposes of water storage for balancing consumption, emergency and fire as well as preparing design pressure. Considering that a significant portion of the
cost of urban water projects is related to tanks and service reservoirs, designers should optimize the reservoir cost as well as its hydraulic and quality performances. Optimization issues in the design and operation phases of water distribution networks mostly focus on pipes or pump's schedule and less on storage capacity, or constraints associated with quality parameters are less
considered. The reason is that water quality shows its importance in the operation stage; therefore, controlling water quality is done in the operation stage with sampling from different parts of the network. In this paper, in order to optimize the performance and quality of the water distribution network, high performance and easy implementation algorithm of NAACO (Non-dominated Archiving Ant Colony Optimization) is developed in Visual C++, which minimizes the cost and maximizes the reliability of storage tanks. To achieve this goal, the Ant Colony Optimization algorithm (ACO) is combined with simulation software (EPANET) which is capable of both hydraulic and quality analyses. Optimal solutions are presented in pareto curves. Application of the proposed approach to the well known test network of Anytown shows the advantages of this methodology. Also, differences in outcomes based on demand
and head driven hydraulic analyses are examined and a relationship between hydraulic reliability of tank and network is proposed. Results show that this methodology can well cover the aspects of the problem and leads to the favorable decisions in the operation of storage tanks in water distribution networks considering quality constraints.