Performance Analysis of Hybrid Genetic Algorithms for the Generalized Assignment Problem

This paper presents three hybrid algorithms, based on genetic algorithm (GA), to solve the generalized assignment problem (GAP) with the objective to minimize the assignment cost under the limitation of the agent capacity. First sequential constructive crossover (SCX) is modified for the problem and then showed competence over one-point crossover (OPX). Our hybrid algorithms use SCX, exchange mutation and three local search algorithms. Experimental results on four sets of benchmark instances from OR-library show the effectiveness of the proposed hybrid algorithms. The proposed algorithms are then compared with bee (BEE) and differential evolution (DE-SK) algorithms. In terms of solution quality as well as computational times, one of our hybrid genetic algorithm (HGA3) outperformed both BEE as well as DE-SK.