چكيده به لاتين
Abstract:
Sewer Networks are among the vital infrastructures in every city and residential area, so their missing will cause severe environmental hazards. Since the construction of such networks is so expensive, any attempt for reducing the costs will result in huge savings. Therefore, many researchers have addressed the sewer network design problem using various methods including non-linear programming, dynamic programming, and meta-heuristic approaches. In a vast number of previous researches, sewer network design problem has been addressed solely to minimize the construction; and system's performance and its reliability in the operation period was neglected or at least rarely addressed.
Due to the importance of layout design in the system's performance and its undeniable effect on the total construction cost, a hybrid method of the Genetic algorithm(GA) and Cellular Automata(CA) is used for the single-objective design of sewer network. Also, a Loop by Loop cutting algorithm is exerted to convert the base looped graph into a dendritic graph. Two approaches were exploited in the single-objective design of sewer networks namely simultaneous design and discrete design. In the discrete design of the sewer network, the layout is optimized based on an objective function which can be reliability or the relative cost. Right after that, the elements of the network including the pipe's size and their buried depth are determined using cellular automata approach. This approach uses a discretization; hence, it reduces the complexity of the sewer design problem. Besides, there is a possibility to design the sewer layout based on different objective functions. On the other hand, the simultaneous design of the sewer network is able to deliver more accurate answers with less total cost.
Two objectives that are considered in this study, reliability, and cost, are in conflict with each other; and any attempt for ameliorating one objective will exacerbate the other. Consequently, the sewer network design is solved using multi-objective approaches. In the first step, the layout problem is presented as a multi-objective problem and it is solved using NSGA-II and some of the Pareto solutions are presented using non-symmetric Nash solutions as practical answers. Moreover, due to the high computation cost for solving the whole problems of sewer network using NSGA-II, a cellular automata algorithm was combined with NSGA-II to reduce the computation time and deliver better depth; likewise, NSGA-II creates initial population and uses non-domination and crowding to render a set of optimal solutions.
Keywords: Sewer Network, Cellular Automata, Layout, Multi-objective Optimization, Reliability
