Development of a cost function of water distribution systems for residential subdivisions

Abstract

This study aims to create a cost function for residential subdivisions based on three key variables: population density, area, and slope. The cost function was developed by minimizing the capital cost of representative residential water distribution networks through a genetic algorithm and a heuristic search method known as the greedy algorithm. To test the proposed cost function and determine if a more efficient design would reduce cost, two subdivisions in Tucson, Arizona, were compared to an equivalent theoretical oblong network. The greedy algorithm required a fraction of the time demanded by the genetic algorithm and arrived at subdivision network costs that were consistently equal to or lower than the best solutions found by the genetic algorithm. Both optimization methods obtained results indicating that area has the greatest effect on cost and that the effect of population density is negligible when dealing with small areas.

Keywords::

• genetic algorithm
• greedy algorithm
• optimization
• residential subdivisions
• heuristics

Acknowledgements

Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.

Funding

This material is based in part upon work supported by the National Science Foundation [grant number 083590].