Identification and modelling of civil formation flight routes based on global flight schedule data

Abstract

This paper presents methods to model potential formation flight routes based on global flight schedule data. The underlying dataset covers a period of one week comprising over 34000 long-haul flights of seven different aircraft types. In a first step, potential formations are identified by a series of different filters that consider parameters of the individual flights such as origin, departure time and initial azimuth angle. In a next step, the origins and departure times of a flight pair are coupled to determine whether the two flights are able to form a formation. If two flights are candidates to form a formation, the flight paths are modeled by a geometric approach. In this approach, all flight paths are generated as great circle segments by an enhanced Fermat point principle for spherical triangles to determine the geometrically shortest possible route for the formation. Based on this Fermat route, the optimal route is derived by taking the departure times of the participating flights into account. The derived optimal route then represents the compromise between the geometric optimum and the input flight schedule. The common formation ruise altitude and Mach number are set according to a performance model and ensure that all aircrafts stay within their performance limits. Surrogate models are then applied to estimate the relative fuel saving of each formation; allowing each flight to be assigned to its most fuel saving formation.