Competitive hub location problem: Model and solution approaches

•We study the hub location problem faced by an airline entering a competitive market in single and multiple allocation network setting.•The problem is modelled as a non-linear integer program.•We propose four different approaches to solve it.•Our best performing method uses Kelley’s cutting plane wi...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Transportation research. Part B: methodological Ročník 146; s. 237 - 261
Hlavní autoři: Tiwari, Richa, Jayaswal, Sachin, Sinha, Ankur
Médium: Journal Article
Jazyk:angličtina
Vydáno: Oxford Elsevier Ltd 01.04.2021
Elsevier Science Ltd
Témata:
Airlines
Algorithms
Competition
Cutting
Datasets
Exact solution methods
Hub and spoke networks
Market shares
Markets
Mixed integer
Non-linear program
Site selection
Competition
Non-linear program
Exact solution methods
Hub and spoke networks
ISSN:0191-2615, 1879-2367
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:•We study the hub location problem faced by an airline entering a competitive market in single and multiple allocation network setting.•The problem is modelled as a non-linear integer program.•We propose four different approaches to solve it.•Our best performing method uses Kelley’s cutting plane within Lagrangian relaxation.•We are able to solve instances upto 50 nodes from AP data-set within 120 and 10 min of CPU time for single and multiple allocation settings, respectively. In this paper, we study the hub location problem of an airline that wants to set up its hub and spoke network, in order to maximize its market share in a competitive market. The market share is maximized under the assumption that customers choose amongst competing airlines on the basis of utility provided by the respective airlines. We provide model formulations for the airline’s problem for two alternate network settings: one in the multiple allocation setting and another in the single allocation setting. Both these formulations are non-linear integer programs, which are intractable for most of the off-the-shelf commercial solvers. We propose two alternate approaches for each of the formulations to solve them optimally. The first among them is based on a mixed integer second order conic program reformulation, and the second uses Kelley’s cutting plane method within Lagrangian relaxation. On the basis of extensive numerical tests on well-known data-sets (CAB and AP), we conclude that the Kelley’s cutting plane within Lagrangian relaxation is computationally the best for both the single and multiple allocation settings, especially for large instances. We are able to solve instances upto 50 nodes from AP data-set within 120 and 10 minutes of CPU time for single and multiple allocation settings, respectively, which were unsolved by mixed integer second order cone based reformulation or Kelley’s cutting plane algorithm in the maximum allowed CPU time (3 hours for single allocation and 1 hour for multiple allocation).
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0191-2615
1879-2367
DOI:10.1016/j.trb.2021.01.012