Abstract:
The hub location decision is a long term investment and any changes in it take
considerable time and money. In real situations, some parameters are
uncertain hence, deterministic models cannot be more efficient. The ability of
two-stage stochastic programming is to make a long-term decision by
considering effects of it in short term decisions simultaneously. In the twostage
stochastic programming for hub location problems, the location is
decided in the first stage and optimal flow allocations are determined in the
second stage. In this paper, the two-stage stochastic programming is described
and then a practical stochastic model is employed for determining hub
locations in the Iranian aviation. Also, a survey of the model under fuel
subsidy omission is conducted using extended two-stage stochastic
programming. Demand and the cost of resources (fuel) are considered
uncertain in this study. The results show Tehran, Mashhad, Isfahan, Shiraz
and Yazd can be hubs in the air network of Iran
Machine summary:
J. Humanities (2012) Vol. 19 (4): (135-154 ) Airline Hub-Median Network Design by an Extended Two-Stage Stochastic Programming Method: A Case Study Mahdi Bashiri1, Aida Omidvar2*, Reza Tavakkoli-Moghaddam3 Received: 2011/10/4 Accepted: 2012/4/14 Abstract The hub location decision is a long term investment and any changes in it take considerable time and money.
Yang proposed a two- stage stochastic model for air freight hub location with uncertain demand (Yang, 2009: 4424-4430).
3. Two-Stage Stochastic Model (Yang’s Model) for Hub-Median Network Usually, studies in hub location problems consider three assumptions: 1) using a discount factor (α) in inter-hub connections for economies of scale and it is a number in unit interval.
The model is: (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) Airline Hub-Median Network Design by an Extended … Constraint (10) ensures that all demands from each origin are transported to its destinations.
proved that stochastic uncapacitated hub location problems with uncertain dependent transportation costs (changing the price of resources such as fuel) are equivalent to the associated expected value problem (Contreras, 2011:518-520).
Table 4: Numerical Results of Model for 20 Nodes in Iran Aviation Hub Network Design Case Study Stochastic Value Deterministic Value Expected Value Total cost ($) 242976101.
3 Tehran, Mashhad, Isfahan, Tehran, Mashhad, Isfahan, Tehran, Mashhad, Isfahan Hub location Shiraz, Yazd Shiraz, Yazd Shiraz, Yazd Non-stop (H, M, L) (50, 54, 54) 54 54 Hub-stop (H, M, L) (330, 326, 326) 326 326 According to the results of stochastic model, the proposed routes are different in three scenarios and the number of direct and hub stop routes are different in high, low and medium demand level.