Font Size: a A A

Research On Leakage Localization And Control Of Municipal Water Supply Network

Posted on:2007-10-09Degree:DoctorType:Dissertation
Country:ChinaCandidate:X LiFull Text:PDF
GTID:1102360212489290Subject:Environmental Engineering
Abstract/Summary:PDF Full Text Request
The research is mainly about leakage detection, localization and leakage control in municipal water supply network. In this paper, the bursts by hydrant flushing are simulated, and the leakage localization model and leakage control model are established to reduce water loss effectively and advance the economic benefit and social benefit of water companies.First of all, from the point of water network leakage detection methods and the influencing factors, the existing leakage detection methods are introduced, which pointed out the shortcomings and put forward new leakage detection ideas according to the actual water networks of our country.Secondly, because the known historical data are imperfect and the data quantity is less, the method of burst simulation is addressed to solve the problem. Then, setting the actual pipe network of some northern city as example, a set of water hydrants across the zone are chosen for burst simulation, and the experiment process, experimental methods and correlative rules are presented, consequently providing data support for establishing an accurate leakage detection model.Thirdly, owing to less monitoring nodes in water network of big city, the collected state variables are limited and can't effectively reflect the operating state of the network, so a BP state estimation model based on Genetic Algorithm (GA) is established, which inquire into the unknown information according to the known monitoring information and make the leakage localization model applied widely.Fourthly, combining the experimental data with the historical data of the leakage accidents, the three models are put up: the leakage localization statistical model based on the Probabilistic Neural Network (PNN), the leakage localization theoretical model based on a Bayesian probabilistic framework, and the theoretical model based on Cluster Analysis and Pattern Recognition. With the exception of this, the applicability of each model is mentioned.In the end, based on the other aspects of water network leakage control-accidents timely repairing and active prevention of accidents, the valves scheduling model in the existing of burst accident and pressure control model through optimizing automatic control valves settings in the normal network operating state are set up. Thus, the network leakage is roundly controlled from the aspect of systems engineering.
Keywords/Search Tags:water supply network, leakage detection and localization, state estimation, auto-adjusting valve, Bayesian theorem, leakage control
PDF Full Text Request
Related items