Experimental analysis of hybrid genetic algorithm for the grey pattern quadratic assignment problem

Abstract

In this paper, we present the results of the extensive computational experiments with the hybrid genetic algorithm (HGA) for solving the grey pattern quadratic assignment problem (GP-QAP). The experiments are on the basis of the component-based methodology where the important algorithmic ingredients (features) of HGA are chosen and carefully examined. The following components were investigated: initial population, selection of parents, crossover procedures, number of offspring per generation, local improvement, replacement of population, population restart. The obtained results of the conducted experiments demonstrate how the methodical redesign (reconfiguration) of particular components improves the overall performance of the hybrid genetic algorithm.