Abstract
Sectorization is a wide used technique that reduces the complexities associated with the analysis and management of drinking water networks. The districts resulting from sectorization are expected to meet certain performance criteria, such as the minimum number of interventions, uniformity in pressures, and similarity in demands, among others. Developing a methodology for optimal district design that meets some of these requirements is an significant challenge. In this work, the advantages in the use of three different criteria are compared. To this, a two-stage approach is presented. The first stage, the communities detection, is based on a Louvain-type algorithm for the maximization of modularity in weighted networks. The second stage, the physical division of the system, is posed as a two-objective optimization problem using a simulated annealing algorithm. The first of these objectives is the number of isolation valves, while for the second objective three different indices are studied: (a) Loss off resilience,(b) Gini coefficient and (c) Standard deviation, the last two are defined to quantify pressures uniformity. The methodology is applied to real study case: a medium-sized network.
Translated title of the contribution | Comparison of indicators for optimal design of District Metered Areas in Water Distribution Networks |
---|---|
Original language | Spanish |
Pages (from-to) | 1397-1406 |
Number of pages | 10 |
Journal | Mecanica Computacional |
Volume | 38 |
Issue number | 42 |
Publication status | Published - 2021 |
Externally published | Yes |
Keywords
- Sectorization
- optimization
- simulated annealing
- Gini coefficient