Efficient methods for including uncertainty and multiple objectives in water resources management models using genetic algorithms

B. S. Minsker, B. Padera, J. B. Smalley

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Incorporating uncertainty and multiple objectives into water resources management models can be a significant challenge. Most existing methods require substantial computational effort that can be prohibitive for large-scale modeling. In this paper, two new genetic algorithm methods are presented that allow for more efficient solution of these types of problems than was previously possible. Noisy genetic algorithms use Monte Carlo-type sampling to sample from a noisy fitness function (objective function). Substantially fewer samples are required to obtain good solutions than with traditional Monte Carlo sampling, so the method is much more efficient than constraint stacking methods that have been used in water resources management models previously. Nondominated sorted genetic algorithms can efficiently identify the entire trade-off surface among multiple objectives in a single model run. Applications of both methods to risk-based corrective action design for contaminated groundwater are presented.

Original languageEnglish (US)
Title of host publicationComputational methods in water resources - Volume 1 - Computational methods for subsurface flow and transport
EditorsL.R. Bentley, J.F. Sykes, C.A. Brebbia, W.G. Gray, G.F. Pinder, L.R. Bentley, J.F. Sykes, C.A. Brebbia, W.G. Gray, G.F. Pinder
PublisherA.A.Balkema
Pages567-572
Number of pages6
ISBN (Print)9058091244
StatePublished - 2000
EventComputational Methods in Water Resources XIII - Calgary, Canada
Duration: Jun 25 2000Jun 29 2000

Other

OtherComputational Methods in Water Resources XIII
CountryCanada
CityCalgary
Period6/25/006/29/00

ASJC Scopus subject areas

  • Earth and Planetary Sciences(all)
  • Engineering(all)
  • Environmental Science(all)

Fingerprint Dive into the research topics of 'Efficient methods for including uncertainty and multiple objectives in water resources management models using genetic algorithms'. Together they form a unique fingerprint.

Cite this