Study On Leakage Localization Of Water Distribution Network Processing Off-line Flow Data

Today, the leakage of water supply pipe network is a unavoidable problem, every city is in the rapid expansion of urbanization, leakage is becoming more and more serious. The premise for controlling leakage thoroughly is finding the point of leakage. All over the world, many experts and scholars have also conducted research on the positioning technology of leakage in great quantities, including checking method based on the model, method of identification based on Bayesian theory, neural network, etc. I found that the SCADA monitoring data in many domestic urban water supply network has not been fully used through field investigation. Generally, the only function is the daily dispatch of the water supply network and the basis of a department for examination and assessment of economic benefits, and then will be shelved in the database, after a period of time, it wound have been removed, making the data with useful information lose value and the traditional approach cannot keep up with the pace of era of data. So, the paper put forward a kind of leak location method based on processing data of off-line flow, the theory for partition of controlling leakage could provide a decision-support. Therefore, it could effectively control the leakage of water supply network that combining the achievement of the paper with theory based on partition. The research content mainly divided into the following:1. Building the mathematical modeling on example of water supply network, setting up a variety of leakage working conditions, obtaining flow data under different working conditions. The purpose of building mathematical modeling on example of water supply network is to explore the method of processing data on flow, because it is impossible for installing flow meter in each section of actual water supply network. The example network could set a variety of different leakage working conditions in advance, and combining data that come from different leakage working conditions. It could improve the breadth and depth in research. The ultimate goal of these work is to obtain all the pipes flow data in different working conditions, and then dealing with different combinational data to verify the feasibility of the research achievement;2. According to the combination data from example water supply network in flow, processing data research is to forecast the general region of the leakage from example water supply network and achieving the desired project objectives;3. Exploring a kind of outliers identification algorithm of water supply network flow data, and building a mathematical modeling for identifying flow data outliers from equipment in the actual water supply network, which can identify outliers in the data processing. All of these is to get clean data and ensure the feasibility and effectiveness of the data processing achievement;4. It could get clean flow data that identifying the outliers before processing the actual flow data of water supply network. Process the clean flow data and predict the possible leakage area of actual water supply network. It could be faster and better that improving the state of water supply network leakage on Combining with the theory that based on distribution of controlling leakage.