Polytechnic University of Madrid
Modelling Taxi Planning
This work introduces taxi planning optimization (TPO) as a methodology to guide airport surface management operations. The optimization model represents competing aircraft using limited ground resources. TPO improves aircraft taxiing routes and their schedule in situations of congestion, minimizing overall taxiing time (TT), and helping taxi planners to meet prespecified goals such as compliance with take-off windows, TT limits, and trajectory conflicts. By considering all simultaneous trajectories during a given planning horizon, TPO’s estimation of TT from the stand to the runways improves over current planning methods. The operational optimization model is a large-scale space-time multi-commodity network with capacity constraints. In addition to its natural use as a real-time taxi planning tool, several TPO variants can be used for design purposes, such as expansion of new infrastructure. TPO is demonstrated using Madrid-Barajas as test airport.
Arrival management (AM), departure management (DM), and gate management (GM) are considered as the primary management tasks in the operation of an airport. AM plans the arrival sequence for landing air- craft in a given time horizon. DM gives the push-back orders to departing aircraft at the stands and establishes ‘calculated’ take-off time windows (CTOTs). GM assigns stands to arriving aircraft. The success of these tasks is closely related to the efficient operation of the airport taxiways, which is commonly known as the ‘taxi planning’ problem. This work introduces taxi planning optimization (TPO) as a methodology to guide surface management operations.
Aircraft taxiing congestion between stands and run- ways represents a major challenge for airport architects, aircraft schedule planners, and real-time taxiing operators. Congestion is typically caused by an inadequate ground infrastructure at the airport to meet the needs for flight movement. Major hubs suffer aircraft delays on the ground which are sometimes aggravated by low visibility conditions caused by rain, fog, or other contingencies. Aircraft taxiing operations along with departure and AM are also critical due to security reasons.
In the context of airport operations management, TPO is not intended to operate as a stand-alone tool. In contrast, it must be coordinated with DM tools (for departing traffic), with AM and GM tools (for arriving traffic) and, of course, with the actual taxi planner of the airport, known as the aircraft ground controller (AGC). Currently, each activity is modelled separately and then coordinated with all other activities. Updated information, such as a delay in the embarkation process, arises frequently during daily airport operations, especially for outbound traffic. Thus, TPO must be flexible to accommodate changing inputs, while being consistent regarding routes and schedules already delivered from past executions. In this dynamic context, an AGC’s requirement is that TPO’s execution time does not exceed a few minutes. Any flight which may use the taxiways within the incumbent planning horizon must be considered in that TPO run.
These researches have been published in four papers about “Taxi Planning” mathematical models. They have received about 200 citations, Google Academic.
Marín, A. Airport management: taxi planning. Ann.Oper. Res., 2006, 143, 189–200.
Marín, A. and Codina E. Network design: taxi planning. Ann. Oper. Res., 2008, 157, 135–151.
Marín, A. and Salmerón, J. Taxi planner optimization: a management tool. Proc. IMechE Vol. 222 Part G: J. Aerospace Engineering, 2008, 1055-1066.
Marín, A. Airport taxi planning: Lagrangian decomposition. J. Adv. Transp. (2011), DOI: 10.1002/atr.175.
During more than 10 years, we have been researching the mathematical models in “Taxi Planning”. You can follow this work in the papers attached. Now, we continue investigating, so we have two new papers to be published.