Library of codes, instances and data sets for the Aircraft Landing Problem (ALP)

	Name	Last modified	Size	Description
	Parent Directory		-
	ikli_instances.zip	2021-03-19 09:29	9.9K
	ikli_datasets.zip	2020-10-16 11:22	10K
	ikli_codes.zip	2020-10-16 10:27	449K

• ikli-datasets.zip

  : contains four datasets for the ALP, constructed from two traffic days on Paris-Orly airport, obtained from the OpenSky Network, one in July 2018, and one in April 2019. The four datasets come from merging these two traffic days and artificially adding Light and Heavy aircraft.
  For each aircraft, the following information is provided:
  • The model of the aircraft (mdl). It represents the ICAO (International Civil Aviation Organization) type of the aircraft (e.g.,

    A320, B737

    , ...)
  • The Wake-Vortex category (category) of the aircraft, that can be Light, Heavy or Medium.
  • The Scheduled Time of Arrival (sta) in HH :MM :SS and in seconds (sta_s). It represents the scheduled time of landing, as stated in the flight plan. The sta is used as a target time in optimization models that take into account airline companies' preferences, which consist in maximizing punctuality by minimizing the deviations from these target times.
  • The Actual Time of Arrival (ata) in HH :MM :SS and in seconds (ata_s). It represents the actual landing time of the aircraft. For aircraft that were artificially added to our data, we assigned artificial ata and ata_s, which are equal to the sta and sta_s (respectively) of these aircraft. The ata and ata_s are provided for information purposes only and are not used in the mathematical models.
  • We also provide realistic delay cost values for each aircraft, based on the two works: Cook and Tanner and Cook et al.. The cost function is a piece-wise linear function, with the following delay break-points: 300, 900, 1800, and 3600 seconds.
    ☞ In the files

    ikli_datasets.zip

    and

    ikli_instances.zip

    , the columns

    cost_300

    ,

    cost_900

    ,

    cost_1800

    , and

    cost_3600

    represent the delay cost slope (euros/second) for 0 to 300 seconds, 300 to 900 seconds, 900 to 1800 seconds, and 1800 to 3600 seconds of delay respectively.
    ☞ To extrapolate the delay cost for aircraft models that are not considered in the works Cook and Tanner and Cook et al., we performed a simple linear regression. Details about this regression are provided in a Jupyter Notebook named

    ikli_DelayCost_regression.ipynb

    , that can be found in the directory

    ikli_codes.zip.




• ikli-instances.zip

  : contains 12 files of test instances generated from these datasets, by simply considering the first N lines (N∈ {30,40,50}) of each of the four datasets. These instances are used in the paper: S. Ikli, C. Mancel, M. Mongeau, X. Olive and E. Rachelson, The aircraft runway scheduling problem: A survey. Computers & Operations Research, volume 132 (20 pages), 2021. These instances together with the data sets are described in the same paper.
  ☞ Safety separation between landings can be easily computed using the Wake-Vortex category (provided in our data) of the leading and the following aircraft, and the 3x3 separation matrix of the ICAO, that can be found here
  ☞ The time-windows usually considered in the ALP can be computed as proposed in this work.




• ikli-codes.zip

  : contains a Jupyter Notebook named

  ikli_DelayCost_regression.ipynb

  , (with two figures

  full-cost.png

  and

  primal-cost.png

  ) that explains the details about the above-mentioned regression.

  ikli-codes.zip

  contains also one python file with the implementation of the three below-mentioned Mixed-Integer Programming (MIP) models from the literature of the ALP. Results of these implementations are also used in our paper The aircraft runway scheduling problem: A survey. Computers & Operations Research, volume 132 (20 pages), 2021.
  • Beasley: J. E. Beasley, M. Krishnamoorthy, Y. M. Sharaiha, D. Abramson, Scheduling aircraft landings-The static case, Transportation Science 34 (2000) 180-197.
  • Salehipour: A. Salehipour, M. Modarres, L. Moslemi Naeni, An efficient hybrid metaheuristic for aircraft landing problem, Computers & Operations Research 40 (2013) 207-213.
  • Furini: F. Furini, M. P. Kidd, C. A. Persiani, P. Toth, Improved rolling horizon approaches to the aircraft sequencing problem, Journal of Scheduling 18 (2015) 435-447.
  ☞The three MIP formulations are implemented using DoCplex, the Python API of the CPLEX solver, version 12.8.

ikli-instances.zip