Optimal Design Of The City Water-Supply System Based On ANN And GA

The water-supply ductwork is a significant component of the city water-supply system, primarily including three projects of gaining water, cleaning water and water-supply ductwork. The water-supply ductwork is the largest sub-system to invest in the whole city water-supply system, about 50～80%. The design of the water-supply ductwork has a direct effect on the investment of the water-supply system. With regard to the optimal design of the water-supply ductwork, numerous scholars at domestic or abroad have done much extensive research and put forward homologous design methods. But there are always some faults in the actual running. Therefore, a new method of pipeline distribution is attempted to bring forward and some ameliorations about pipe diameter combination optimization are made in order to meet desire. Three tasks are completed in this paper as following,1. The content and significance about the optimal design of water-supply system are expatiated. Some optimal methods which have been advanced are introduced briefly. The function, components and modeled methods of the water-supply system are summarized. The mathematical model of optimal design of water-supply system is set up. The parameters of the target function in the model are analyzed and corrected.2. The basic theory of Hopfield Neural Network (HNN) and its applications in solving compounding optimizing problems such as MTSP are introduced. The optimal distribution of water-supply ductwork by applying HNN based on MTSP is researched in detail and the shortest distance of pipeline is gained.3. The basic principle and operating process of Genetic Algorithms(GA) and its possible analysis in the pipe diameter combination optimization design of water-supplysystem are researched. A program chart of GA's solving the pipe diameter combination optimization problem of water-supply system is presented in the end.