Leak Detection of Municipal Water Supply Network Based on the Cluster-Analysis and Fuzzy Pattern Recognition

Considering a Supervisory Control And Data Acquisition (SCADA) system has been installed in many municipal water supply networks, the water network scale mostly is big and the district meter area (DMA) have not been realized in China, a leak detection model based on the cluster-analysis and fuzzy pattern recognition theorem is introduced. First, computer simulation of pipe leak is realized, and a database of leak simulation on the water network established by Delphi combining with MSSQL programming technology. Then "virtual partitions" is completed by cluster analysis of multivariate statistical. And then, fuzzy recognition is used to identify which virtual region the leak pipe belongs to, which reduced the scope for the leak detection greatly and decreased the difficulty for the detection. At last, by the model examination from an actual water network, the model is feasible and the result is satisfying.