Application of heuristic algorithms on groundwater pumping source identification problems

This study investigates the application of heuristic algorithms on groundwater pumping source identification problem. Unknown groundwater extraction site is located by analyzing the hydraulic heads of the observation wells. Genetic algorithm (GA), particle swarm optimization (PSO), mutated PSO (MUPSO), their hybrid models GA-MUPSO, and groundwater simulation models are employed to solve the problems afore mentioned. The results indicate that all optimization models developed here can successfully locate the pumping source. Especially the GA-MUPSO hybrid model illustrated the best performance and efficiency, not only located the pumping source with a 100% success rate, but also required a far less CPU time than GA and PSO.