Bursts Detection in Water Distribution Networks Based on Real-Time State Simulation

Abstract

Based on the improved weighted-least-square model and fuzzy similarity ratio method, a methodology is proposed to detect pipe bursts in real-time. When SCADA data is obtained DFP algorithm is used to get the real network state. Then the real values of burst characteristics are computed. And the hypothetical values assuming each pipe as the accident pipe are calculated for comparison. The fuzzy similarity ratio method is used to judge whether there is a pipe burst. If there is, the hypothetical value that is most similar to the real value is the accidental state and the corresponding assumed break is the burst location. According to the methodology a software system is developed with Delphi 7 for verification. The running results of a designed network show that the methodology is reliable and its detection accuracy is over 45%.