The Study And Practice On Intelligent Control Technology For Leakage In Water Supply System

City water supply system is the lifeblood of a city, in recent years, with problem of the pipe network leakage has become more prominent and the research for the leakage control has been even more thorough, leakage control begins to develope from the traditional control methods to the direction of intelligent, digital control methods.For the current domestic and foreign leakage situation, this paper summarizes and analyzes the existing basic form of pipe network leakage and the top ten factors that may affect leakage, in order to control leakage effectively, we must study new methods integrated with the traditional control methods. The domestic and foreign progress of leakage detection technology and theory research in prensent days is summarized, finally make a feasible study for the project based on the advantages and disadvantages of intelligent control methods.Pipe network modeling is the basic of leakage intelligent control. The basic ideas and processes of pipe network modeling are introduced based on pipe network Micro-hydraulic model theory, the layout optimization method of pressure monitoring points in the pipe network modeling process is studied focusly, finally the function modules of the Maanshan water pipe network online optimization scheduling system are introduced.Program of gray model compiled on C + + language is used to predict the leakage frequency of the Maanshan pipe network within 12 years, the maximum relative error is 5.18%; to predict the occurrence time of leakage on a pipe seleceted of the city, the error is in the acceptable range; respectively using the three exponential smoothing model and gray neural network model (GRBF) to forecast the the pipe network leakage between 2003～2004, the results show that GRBF model is more accurate than three exponential smoothing model. Concluded that compared with the accuracy of leakage prediction simply relying on datas, the accuracy of leakage prediction considering many influtial factors is improved significantly.Above all, contrary to the characteristics of Maanshan water supply system, the city's pipe network leakage safety evaluation system is established, which is divided into 6 guideline layers and 24 indicator layers. Then AHP combined with fuzzy evaluation is used to carry out a comprehensive safety evaluation of the leakage on the city's pipe network, the results show that the city's pipe network is on "generally safe" level, and according to the results , various leakage causes and its impact are analyzed to verify that the evaluation model is reasonable. At last, the function of each module of the digital management system and the needs to improve and perfect the contents of the system are summarized.