TANTao,ZHANG Miao,JIANGYunlong,et al.Topological Distance?based Clustering Method for Water Supply Network Partitioning[J].China Water & Wastewater,2026,42(4):10-16.
Topological Distance?based Clustering Method for Water Supply Network Partitioning
China Water & Wastewater[ISSN:1000-4062/CN:12-1073/TU]
volume:
第42卷
Number:
第4期
Page:
10-16
Column:
Date of publication:
2026-02-17
- Keywords:
- water supply network; pressure reducing valve; pumping station; topological distance; K-means clustering method
- Abstract:
- District metering is a method of simplifying pressure management by partitioning the water supply network, which can effectively reduce leakage rates. However, current research often focuses solely on the topological relationship of nodes, ignoring the influence of pressure reducing valves and pumping stations on partitioning outcomes. To address this issue, a topological distance-based clustering method for water supply network partitioning is proposed. This method uses pressure reducing valves and pumping stations as the basis for partition boundaries, thereby enhancing partitioning accuracy. Initially, the water supply network is modeled as an undirected graph to construct the adjacency matrix, and the topological distances between pressure reducing valves and pumping stations are magnified. Subsequently, the topological distance matrix is generated using Floyd’s shortest path algorithm. Finally, nodes are clustered based on the constructed topological distance matrix using the K-means clustering method, thus facilitating the partitioning of the water supply network. The proposed method was applied to three case studies and compared with two traditional methods. The results indicate that the proposed partitioning algorithm based on topological distance clustering, yields more accurate partitioning outcomes while ensuring high modularity and uniformity in the partitioning process.
Last Update:
2026-02-17