The Genetic Algorithm Optimization In The Drainage Network Design Applied Research

With the progress of urbanization and the enlargement of city scale in China, the optimal design on new construction rehabilitation and extension of urban drainage networks poses a key task to the water and wastewater engineers. The rehabilitation and extension project of combined sewer system in certain Economic and Technological Development Zone is taken as a case study in this thesis. Genetic Algorithms (GA), which were applied widely in engineering in recent years, is used in the design of this new construction rehabilitation and extension project. Gas is a strong robust global searching algorithm. It is not based on objective function gradient estimates, but directly targets at the code in certain form of the variable. They may be used to solve nonlinear problems such as the drainage networks optimization and reach discrete pipe diameters. After comparing the simulation experiment results of different programs where varies of operation parameters are combined, the program, which consists of population size 20, character coding, elitism strategy, adaptive crossover and mutation rates, could converge the most rapidly in the optimal solution. Finally, the designed optimal sewer system is checked on its hydraulic reliability.