Research On Simulated Annealing Genetic Algorithm For The Optimization Of Urban Water Network

Urban water supply system is one part of important fundamental facilities of model city. And it is a mark of civilization and economy and modernization level. Urban water distribution network influence obviously not only the total cost of water supply engineering and energy consumption during management but also water quality on network layout and pipes'size.Today it is thought much by researchers in domestic and overseas that using genetic algorithm to optimize the network. How to deal with constraints is one important part of focuses which are very important in genetic algorithm methods to optimization constraint functions. It is a key to how to balance the pressure from between the optimization of objective function and the satisfaction of constraints. Penalty function method is mainly used for optimization of constraint functions. But it is difficult to design suitable penalty factor to balance the pressure. Practically the shortcomings include the bad astringency and the premature solution of optimization and so on when using genetic algorithm to optimize the network. So the method can't be used really in optimization problems.Simulated Annealing is an universal algorithm and it is easily accomplished. And it can be effectively avoid premature solution. But it shortcomings include long computer time and low efficiency when using Simulated Annealing to optimize the network. Nevertheless Simulated Annealing has been used widely in engineering.This study is successfully used ID (Infeasibility Degree) to deal with constraints and avoid the difficulty how to design suitable penalty factor when using self-adaptive Simulated Annealing Genetic Algorithm for optimization constraint functions based on the analyzing and comparing some optimal methods. Author has applied successfully this method to Engineering examples which are optimal design of water supply networks for test it. The test has showed the optimal algorithm is better than traditional genetic algorithm on astringency velocity and solution quality.In the end, author has analyzed and explored on capabilities of Simulated Annealing Genetic Algorithm method and given the bounds of parameters in Simulated Annealing Genetic Algorithm as advices.