Robust optimal design of a tree-based water distribution network with intermittent demand

This paper discusses the design of a tree-shaped water distribution system for small, dispersed rural communities. It revisits the topic that was discussed in the literature and is nowadays implemented in the field. It proposes a new approach to pipe selection based on robust optimization to account...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:European journal of operational research Ročník 319; číslo 3; s. 834 - 844
Hlavní autoři: Babonneau, F., Gilbert, D., Piller, O., Vial, J.P.
Médium: Journal Article
Jazyk:angličtina
Vydáno: Elsevier B.V 16.12.2024
Elsevier
Témata:
Computer Science
Gravity-driven distribution systems
Modeling and Simulation
Operations Research
Optimal design
Robust optimization
Simulation
Robust optimization
Gravity-driven distribution systems
Simulation
Optimal design
Gravity-driven flow
Distribution systems
ISSN:0377-2217, 1872-6860
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:This paper discusses the design of a tree-shaped water distribution system for small, dispersed rural communities. It revisits the topic that was discussed in the literature and is nowadays implemented in the field. It proposes a new approach to pipe selection based on robust optimization to account for the uncertainty inherent in intermittent demands. It also proposes a fast projected reduced Newton method of calculating stationary flows to test the performance of the networks thus designed by Monte-Carlo simulation. Numerical experiments conducted on real study cases have shown promising results both in terms of quality and performance of the generated robust solutions and in terms of computation time for simulations. •The design of a tree-shaped water distribution under uncertainty is discussed.•A robust optimization-based approach to pipe selection is introduced.•A fast projected reduced Newton method of calculating stationary flows is proposed.•Numerical experiments conducted on real study cases show promising results.
ISSN:0377-2217
1872-6860
DOI:10.1016/j.ejor.2024.07.020