Sensitivity Analysis of a Genetic Algorithm for a Competitive Facility Location Problem

Abstract

The paper describes an application of a multi-objective evolutionary algorithm that finds Pareto-optimal solutions to a competitive facility location problem and reports a sensitivity analysis of the evolutionary algorithm model. A genetic algorithm which uses non-uniform mutation and continuous recombination is developed to find solutions to the bi-objective location problem. Results show that the genetic algorithm can find solutions close to the Pareto-optimal set in polynomial time and results from sensitivity analysis show that the mutation rate has the largest influence on the output among the input factors.