Title page for ETD etd-08282001-160145

Type of Document Master's Thesis
Author Zhu, Xiaomei
URN etd-08282001-160145
Title A Demand Driven Re-fleeting Approach for Aircraft Assignment Under Uncertainty
Degree Master of Science
Department Industrial and Systems Engineering
Advisory Committee
Advisor Name Title
Bish, Ebru K. Committee Co-Chair
Sherali, Hanif D. Committee Co-Chair
Trani, Antonio A. Committee Member
  • Fleet
  • Airline Operations Research
  • Demand Driven Re-fleeting
Date of Defense 2001-07-23
Availability unrestricted
The current airline practice is to assign aircraft capacity to scheduled

flights well in advance of departure. At such an early stage in this process,

the high uncertainty of demand poses a major impediment for airlines to best

match the airplane capacities with the final demand. However, the accuracy of

the demand forecast improves markedly over time, and revisions to the initial

fleet assignment become naturally pertinent when the observed demand

considerably differs from the assigned aircraft capacity. The Demand Driven

Re-fleeting (DDR) approach proposed in this thesis offers a dynamic

re-assignment of aircraft capacity to the flight network, as and when improved

demand forecasts become available, so as to maximize the total revenue.

Because of the need to preserve the initial crew schedule, this re-assignment

approach is limited within a single family of aircraft and to the flights

assigned to this particular family. This restriction significantly reduces the

problem size. As a result, it becomes computationally tractable to include

path level demand information into the DDR model, although the problem size

can then get very large because of the numerous combinations of composing

paths from legs. As an extension, models considering path-class level

differences, day-of-week demand variations, and re-capture effects are also


The DDR model for a single family with path level demand considerations is

formulated as a mixed-integer programming problem. The model's polyhedral

structure is studied to explore ways for tightening its representation and for

deriving certain classes of valid inequalities. Various approaches for

implementing such reformulation techniques are investigated and tested. The

best of these procedures for solving large-scale challenging instances of the

problem turns out to be an integrated approach that uses certain selected

model augmentations and valid inequalities generated via a suitable separation

routine and a partial convex hull construction process. Using this strategy in

concert with properly selected CPLEX options reduces the CPU time by an

average factor of 7.48 over an initial model for a test-bed of problems each

having 200 flights in total. Prompted by this integrated heuristic approach, a

procedure for finding solutions within a prescribed limit of optimality is

suggested. To demonstrate the effectiveness of these developed methodologies,

we also solved two large-scale practical-sized networks that respectively

involve 800 and 1060 flights, and 18196 and 33105 paths in total, with 300 and

396 flights belonging to the designated family. These problems were typically

solved within 6 hours on a SUN Ultra 1 Workstation having 260 MB RAM and a

clock-speed of 167 MHz, with one exception that required 14 hours of CPU time.

This level of computational effort is acceptable considering that such models

are solved at a planning stage in the decision process.

  Filename       Size       Approximate Download Time (Hours:Minutes:Seconds) 
 28.8 Modem   56K Modem   ISDN (64 Kb)   ISDN (128 Kb)   Higher-speed Access 
  meithesisetd.pdf 5.32 Mb 00:24:36 00:12:39 00:11:04 00:05:32 00:00:28

Browse All Available ETDs by ( Author | Department )

dla home
etds imagebase journals news ereserve special collections
virgnia tech home contact dla university libraries

If you have questions or technical problems, please Contact DLA.