Plume containment/cleanup strategy design by genetic algorithm methods
Simulation/Optimization (S/0) modules were applied to develop an optimal pump and treat (P&T) pumping strategy for a Wurtsmith AFB site. The S/0 model incorporates both operations research linear programming and evolutionary optimization genetic algorithm (GA). Differing from a previously reported GA code that treats pumping as a discrete variable (no pumping or pumping at some specified rate), this code treats pumping as a continuous variable. It was used to design the cleanup wells which would reduce contaminant concentrations to acceptable values within 6 years. The linear programming capability of REMAX was then used to determine the least amount of extra pumping needed to achieve plume containment. Results of this optimization approach will be presented.
Peralta, R.C. and A.H. Aly. 1998. Plume containment/cleanup strategy design by artificial neural network/genetic algorithm methods. Invited presentation at the US Dept. of Defense Environmental Security Second Annual Modeling and Simulation Conference, 4 May 1998, Alexandria, VA. (D)