HR: 1340h
AN: H53F-0546    [Abstracts]
TI: Optimizing Groundwater Remediation Designs Under Uncertainty Using Dynamic Surrogate Models
AU: * Yan, S
EM: smyan@uiuc.edu
AF: University of Illinois, 4129 Newmark Lab, MC-250 205 N. Mathews Ave, Urbana, IL 61801 United States
AU: Minsker, B
EM: minsker@uiuc.edu
AF: University of Illinois, 3230 Newmark Lab, MC-250 205 N. Mathews Ave, Urbana, IL 61801 United States
AB: Computational cost is a critical issue for large-scale water resource optimization problems that often involve time-consuming simulation models. This issue is compounded when optimizing under uncertainty, since Monte Carlo simulations are often required to evaluate objective function values over multiple parameter realizations. In order to improve computational efficiency, we propose a soft computing approach, in which the time-consuming numerical models are approximated and replaced by dynamic surrogates embedded within a noisy genetic algorithm (GA) optimization framework. The surrogates are trained to predict the distribution of the objectives online, using Monte Carlo simulation results created during the GA run. The surrogates are then adaptively updated to improve their prediction performance and correct the GA­_s convergence as the search progresses. Latin Hypercube sampling method is used to efficiently sample parameters for the Monte Carlo simulation and the sampling results are archived so that the estimation of the objective function distributions is progressively improved. The GA is modified to incorporate hypothesis tests to produce reliable solutions. The method is applied to two groundwater remediation design case studies, where the primary source of uncertainty stems from hydraulic conductivity values in the aquifers. Our preliminary results show that the technique can lead to reliable and cost-effective solutions with significantly less computational effort.
DE: 0555 Neural networks, fuzzy logic, machine learning
DE: 1829 Groundwater hydrology
DE: 1847 Modeling
DE: 1873 Uncertainty assessment (3275)
DE: 4255 Numerical modeling (0545, 0560)
SC: Hydrology [H]
MN: Fall Meeting 2005