Novel approach to deal with demand volatility on fleet assignment models

Abstract

One of the important applications of operations research in the airline industry is fleet assignment. The problem is posed as an assignment of aircraft capacity to flight legs at the planning level, in general one year before departure date. The fleet assignment problem comes after schedule design and without any influence on it. Even if the schedule is defined by a set of flight legs, the source of revenues for airlines is itineraries, many of which have more than one leg. Existing research is based on the itinerary-based fleet assignment model (IFAM) that captures the network effects. Nevertheless, the difficulty of forecasting itinerary demand prevents the widespread implementation in the airline industry of the IFAM and impacts heavily on its performance. This paper proposes a new model based on itinerary grouping. Our itinerary group fleet assignment model (IGFAM) deals with the difficulties caused by itinerary forecast by replacing them with aggregated demand forecasts. We conduct comparisons between models, considering their respective profit based on real-life demand using a simulation framework. Though the comparison is conservative, it still leads to the new model delivering an advantage in almost all circumstances. The greatest benefits are observed at the highest demand volatility.

Keywords:

• Fleet assignment
• airlines
• clustering
• passenger mix model

Notes

Please note this paper has been re-typeset by Taylor & Francis from the manuscript originally provided to the previous publisher.