Urban Water Distribution Netword Model Research For Optimal Design Of Reconstruction And Extension

Urban water supply system is a set of engineering facilities, which is composed of water intake units, water treatment structures, pumping stations, pipes (water distribution network) and so on. Water supply network is an important part in urban water supply system, for its investment is often accounted for 60% to 80% of the total investment of the entire water supply system project, during the operation and also needs to put into a huge power expense and management and maintenance fees each year.At present the water supply system of each city is of sufficient size in general. Because of pipe network system is running up to a certain number of years, it appears such problems:the low efficiency of water supply, serious secondary pollution of pipe network, high rate of pipe burst and leakage, unscientific operation and management.etc. Therefore, it is necessary to optimize rehabilitation and expansion of the existing water supply network. Optimal design of reconstruction and extension is to tap water supply capacity and transform of the old pipe, to decide how to lay new pipe economic and reasonable, to coordinate the old and new pipe network, which should meet user requirements in the amount of water, water pressure, water quality, safety and reliability, and the lowest cost of reconstruction and extension.Through the research of existing pipe network optimal renovation and expansion model, this paper presents a more comprehensive model which involved in a variety of economic and technical parameters. In this model, system reliability is defined as the node and pipeline reliability; to consider the qualitative constraints of water hammer and water quality; to combine the optimal selection and scheduling of pumps with the optimal expansion design so that the cost of pipeline construction, operation and management is the most economical.Through the analysis and comparison on the common algorithms for water distribution network optimization model such as non-linear programming, dynamic programming, neural networks, genetic algorithms, ant colony algorithm, an improved genetic algorithm basing on an integer code is proposed to solve the non-linear model. During the process, choose the node-formula (a hydraulic calculation method) to calculate and combine with genetic algorithm. This method can use the node pressures calculated directly to determine the level of individual fitness, improve the computing speed and reduce the amount of calculation.Finally, this model is verified by its application on a representative reconstruction and extension of water distribution network optimization engineering example, using improved genetic algorithm. Comparing the optimal results to the results with the traditional design method, it fully illustrates that there are important theoretical and application value of the model and algorithm in engineering practice. Optimizing the selection and scheduling of the pump station based on the optimal design of water supply network can reduce the reconstruction and extension operating costs further and save energy.