The capacitated p-hub median problem with integral constraints: An application to a Chinese air cargo network

Cheng Chang Lin, Jr Yung Lin, Yin Chieh Chen

Research output: Contribution to journalArticlepeer-review

42 Citations (Scopus)

Abstract

Consolidation at hubs in a pure hub-and-spoke network eliminates partial center-to-center direct loads, resulting in savings in transportation costs. In this research, we propose a general capacitated p-hub median model, with economies of scale and integral constraints on the paths. This model requires the selection of a specific p among a set of candidate hubs so that the total cost on the resulting pure capacitated hub-and-spoke network is minimized while simultaneously meeting origin-destination demands, operational capacity and singular path constraints. We explored the problem structure and developed a genetic algorithm using the path for encoding. This algorithm is capable of determining local optimality within less than 0.1% of the Lagrangian relaxation lower bounds on our Chinese air cargo network testing case and has reasonable computational times. The study showed that designating airports with high pickups or deliveries as hubs resulted in a high percentage of origin-destination pairs (ODs) in direct deliveries. Furthermore, the more hubs there are, the higher the direct share and the less likely for double rehandles. Sensitivity analysis on the discount rate showed that the economies of scale on trunk lines of hub-and-spoke networks may have a substantial impact on both the operating costs and the route patterns.

Original languageEnglish
Pages (from-to)2777-2787
Number of pages11
JournalApplied Mathematical Modelling
Volume36
Issue number6
DOIs
Publication statusPublished - 2012 Jun

All Science Journal Classification (ASJC) codes

  • Modelling and Simulation
  • Applied Mathematics

Fingerprint

Dive into the research topics of 'The capacitated p-hub median problem with integral constraints: An application to a Chinese air cargo network'. Together they form a unique fingerprint.

Cite this