Research On Intelligent DMA Zoning Optimization Of Water Supply Network Based On Spectral Clustering Algorithm

With the rapid development of urbanization,the huge economic losses and negative social effects caused by the leakage of the water supply pipe network are also increasing day by day.The leakage of the water supply pipe network has been paid more and more attention by the society,and it is also a big problem that needs to be solved urgently by smart water service..After many years of practice at home and abroad,the water supply network DMA(District Metered Area,divide the water supply network into different independent metering areas)With its advanced management methods and immediate leakage control performance,the District Metered Area method has become one of the leakage control technologies for the water supply network.The purpose of this paper is to research a method,which is more scientific and reasonable DMA partition analyze the principles and procedures of DMA partition,aim at the complex topological structure of the water supply network and the intricately connected pipelines,use the spectral clustering algorithm to connect the nodes and establish the similarity between the nodes.Degree matrix,using spectral clustering algorithm to divide multi-region DMAs(i.e.DMAs).This method provides a scientific basis for the realization of intelligent DMA partitioning of the water supply network,and Partly get rid of the dependence of traditional artificial experience partition on subjective ideas.The specific research work and contributions of this paper are as follows:(1)Acquisition and processing of sample pipe network data sets.Use the literature survey method to query historical data and obtain the data required for example pipe network modeling,including GIS information listed in the node,pipe network topology information,pipe network length and pipe diameter and other information.After obtaining the data,conduct on-site surveys,And correct the wrong information.By dividing,summarizing and sorting out the attributes of the data set,it lays the foundation for the construction of a pipe network hydraulic model on the WaterGEMS platform in the next step.(2)Construct a hydraulic model of the water supply network.This research is based on the actual pipe network data,using the nodes,pipelines,and hydraulic data of the pipe network to build a micro hydraulic model on the WaterGEMS modeling platform,and improve it through manual pre-checking of the pipe network model and accurate calculation checking.The accuracy of the pipe network model is improved.Verify the gap between the model and the actual to analyze its feasibility.(3)Realization of DMA division of water supply network.Aiming at the traditional clustering algorithm for non-convex spherical data distribution sample spatial clustering,it is easy to fall into the local optimal solution.Aiming at the problem of poor clustering effect,this study proposes to use spectral clustering algorithm to divide DMA partitions,and to divide nodes when dividing boundaries.The energy redundancy difference between the nodes is regarded as the similarity between nodes,clustered according to the degree of similarity,and then the natural neighbor concept of nodes is used to reduce the number of DMA boundary pipe sections.Finally,the study finally divides the example pipe network into 23 DMA partitions.The results show that the DMA partition results obtained by using this method can not only reduce the amount of leakage,but also have advantages in improving the operating conditions of the pipe network system.(4)In this research,a hydraulic model of the water supply network in City L was established based on the basic data of the pipe network in the example of City L.Finally,three indicators were used to verify the effect of the DMA partition.The verification indicators included three indicators: leakage rate,pressure uniformity and water reliability.Examples have proved that the water supply network of L city after the DMA optimized partition is more leaky than before the partition.A reduction of 4.11%.It proves the rationality and feasibility of the DMA partition optimization research based on the spectral clustering algorithm,and has good economic benefits.