Multi-objective reliable hub covering location considering customer convenience using NSGA-II

Abstract

In hub location problem the overall network performance is extremely depends on the operation of network elements such as hubs and arcs. Happening disorder in one elements of network (hubs or arcs), may lead to network breakdown or poor service levels. This study presents multi-objective single allocation hub covering location which is by exploring the tradeoff between the total cost and total number of intermediate links between all origins and destinations, offers a scenario ideal from the investor point of view and so convenient from the customer point of view. This study also gives this assurance that the minimum reliability of all routes that formed by hubs and arcs does not less than a specified value (controls the network minimum reliability). The type of coverage that rarely studied in the literature and does not have other type’s deficiency is used. Moreover, this paper presents a non-dominated sorting genetic algorithm-II, which its parameters are tuned by response surface methodology, to find near optimal pareto solutions for the given problem. The reliability of the proposed algorithm, based on three evaluation metrics (number of pareto solutions, spacing metric, diversity metric) is determined. Computational results of Turkish dataset are presented for better illustration of the proposed model.