A heuristic method based on LP relexation to solve the multiple allocation modular hub location problem

Abstract

Purpose: In this research, a modular hub location problem has been investigated where the objective is to reduce the transportation costs in the hub network. The proposed model determines the location of hubs, allocation of the non-hub nodes to the hubs, and the optimal vehicle traffic, i.e., the number of flights or the number of trucks traveling in the network, considering the appropriate capacity for each vehicle. Also, decisions regarding the percentage of the traffic volume sent via multiple network routes are made by the presented model.Methodology: The mathematical model, including the objective function and constraints, is constructed and solved by GAMS software. The effect of different parameters on the results is investigated. Due to long solution times for the MIP model, a heuristic solution method based on LP relaxation of the integer variables is developed for the proposed problem, which is able to obtain near-optimal solutions in less time.Findings: The developed mathematical model is implemented on the air passenger transportation data for the airports of the United States of America, which is known as the CAB data set. The results give the optimal number of hubs, as well as the optimal number of transportation units on each arc of the network, which depend on the capacity of the means of transportation.Originality/Value: In this research, a mixed integer programming model is developed for the multiple allocation modular hub location problem. Numerical experiments are conducted with the use of GAMS software and the results are discussed.

Keywords

• modular hub location
• multiple allocation
• mixed integer programming
• lp relaxation