Author(s): Roberto Del Teso; Elena Gomez; Alvaro Montana; Elvira Estruch-Juan

Linked Author(s):

Keywords: No Keywords

Abstract: The analysis of solutions for designing water distribution networks must encompass various aspects. One aspect to consider is the energy required to supply the system, aiming to achieve efficient and sustainable systems. It is crucial to define a layout that is energetically favorable, adapting to the morphology of the area and user requirements. Another key aspect is the implementation cost of the network. This work proposes a methodology to identify a set of layouts deemed energetically favorable, requiring less head energy, exhibiting lower energy excesses at nodes, and incurring lower total costs, including energy costs during operation. The proposed methodology employs clustering, specifically utilizing K-means and hierarchical clustering algorithms. These algorithms are valuable for identifying patterns or structures without the need for labels or external information, enabling the grouping of data based on similarities. By leveraging graph theory, potential routes for pipe installation are determined. This involves considering both the available paths on which the network can be laid out and the base topology generated based on the initial sectors (clusters). This base topology corresponds to an undirected graph, serving as the foundation for generating all possible layouts for the water distribution network.

DOI:

Year: 2024