A hybrid ant colony optimization algorithm for solving facility layout problems formulated as quadratic assignment problems

摘要

Purpose - This paper aims to describe a new hybrid ant colony optimization (ACO) algorithmrndeveloped to solve facility layout problems (FLPs) formulated as quadratic assignment problemsrn(QAPs).rnDesign/methodology/approach - A hybrid ACO algorithm which combines max-min ant systemrn(MMAS) (I.e. a variant of ACO) with genetic algorithm (GA) has been developed. The hybridrnalgorithm is further improved with the use of a novel minimum pheromone threshold strategyrn(MPTS).rnFindings - The hybrid algorithm shows satisfactory results in the experimental evaluation due tornthe synergy and collaboration between MMAS and GA. The results also show that the use of MPTSrnhelps them to achieve such performance, by promoting search diversification.rnResearch limitations/implications - The experimental evaluation presented emphasizes more onrnthe search performance or pattern of the hybrid algorithm. Detailed computational work could revealrnother strengths of the algorithm.rnPractical implications - The developmental work presented in this paper could be used byrnresearchers and practitioners to solve QAPs. Its use may also be expanded to solve otherrncombinatorial optimization and engineering problems.rnOriginality/value - This paper provides useful insights into the development of a hybrid ACOrnalgorithm that combines MMAS with GA for solving QAPs.