Water Supply Network Partitioning Based On Weighted Spectral Clustering

Abstract

Water Network Partitioning (WNP) in District Meter Area (DMA), obtained inserting remote control valves and flow meters in water supply systems, allows simplifying the water balance and pressure control in order to reduce water leakage and to improve water quality protection. Traditionally, the WNP is based on empirical suggestions and on trial and error approaches used with hydraulic simulation software, difficult to apply to large networks. Recently, some heuristic procedures, based on graph and network theory, have shown that it is possible to find optimal solutions in terms of number, shape and dimension of DMAs. In this paper, spectral clustering theory was used to define the water districts, taking into account the spatial and hydraulic constraints, through weight matrices. A comparison between different spectral clustering methods was achieved on a real water network measuring some energy performance indices, in order to identify the optimal water network partitioning.