A matheuristic approach to preference aware air traffic flow management

The main objective of Air Traffic Flow Management (ATFM) is to assign 4D trajectories to flights to guarantee both safe operations and efficient use of airspace. In this talk, we present a new approach to ATFM that explicitly takes AUs routes’ preference into account. A set of relevant trajectories is extracted from air traffic data repositories, using data analytics to learn the preference of each flight for each trajectory. The information feeds an Integer Linear Programming (ILP) model where variables assign a trajectory and a ground delay to each flight. For real-size instances, the number of trajectories leads to prohibitively large models and we propose a heuristic that, at each iteration, solves the ILP model restricted to a suitable subset of variables determined through machine learning techniques. The randomized selection comes from a tree classifier that considers features related to candidate variables, as well as to current solution and search state, and is preliminarily trained on reduced size instances. Computational results, compared to a heuristic based on column generation, show the ability of the proposed method to effectively solve realistic instances and sensibly reduce running times while preserving the quality of the solutions. [1] Estes, A.S, Ball, M.O.: Alternative Resource Allocation Mechanisms for the Collaborative Trajectory Options Program (CTOP). In Thirteenth USA/Europe AirTraffic Management Research and Development Seminar, 1-9, (2019) [2] Lancia, C., De Giovanni, L., Lulli, G.: Data analytics for trajectory selection and preference-model extrapolation in the European airspace. In: The OR2018 Proceedings. Springer, 7 p., (2018)