| Abstract |
Highlights: What are the main findings? A method for clustering urban water supply consumers based on their geographic location and water consumption levels has been developed, allowing for the identification of key areas with high water demand and consideration of spatial distribution characteristics on the map. The effectiveness of applying the agglomerative clustering method has been demonstrated, which, combined with the silhouette coefficient, allows for selecting the optimal data partition strategy and distributing the territorial division of consumers across service zones. What are the implications of the main findings? The proposed geospatial clustering method allows for planning the placement of pressure sensors, taking into account the distance of consumers from water sources, consumption centers, and their cluster membership, which creates a foundation for solving hydraulic modeling tasks. The results of the study are aimed at addressing the task of determining the magnitude and influence zones of pressures generated by water sources on the key points of the urban water supply system, in order to optimize the operation modes of pump stations. For large cities with developing infrastructures, optimising water supply systems plays a crucial role. However, without a clear understanding of the network structure and water consumption patterns, addressing these challenges becomes significantly more complex. This paper proposes a methodology for geospatial data analysis aimed at solving two key tasks. The first is the delineation of service zones for infrastructure objects to enhance system manageability. The second involves the development of an approach for the optimal placement of devices to collect and transmit hydraulic network parameters, ensuring their alignment with both water supply sources and serviced areas. The study focuses on data from the water supply network of a city with a population exceeding half a million people, where hierarchical clustering using Ward’s method was applied to analyse territorial distribution. Four territorial clusters were identified, each characterised by unique attributes reflecting consumer concentration and water consumption volumes. The cluster boundaries were compared with the existing service scheme of the system, confirming their alignment with real infrastructure. The quality of clustering was further evaluated using the silhouette coefficient, which validated the high accuracy and reliability of the chosen approach. The paper demonstrates the effectiveness of cluster boundary visualisation for assessing the uniform distribution of pressure sensors within the urban water supply network. The results of the study show that integrating geographic data with water consumption information not only facilitates effective infrastructure planning and resource allocation but also lays the foundation for the digitalization of the hydraulic network, a critical component of sustainable development in modern smart cities. © 2025 by the authors. |
