A Distributed Framework for Traffic Flow Management in the Presence of Unmanned Aircraft

Abstract

The integration of unmanned aircraft systems (UAS) into the airspace system is a key challenge facing air traffic management today. An important aspect of this challenge is how to determine and manage 4-dimensional trajectories for both manned and unmanned aircraft, and how to appropriately allocate resources among different aircraft. An integrated approach requires solving the traditional Air Traffic Flow Management (ATFM) problem to balance the capacity and demand of airport and airspace resources, but at a significantly larger scale. In doing so, aircraft connectivity constraints of commercial flights must be satisfied. In addition to these and the resource capacity constraints, geofencing constraints for unmanned aircraft that keep them within or outside a certain region of the airspace, must also be incorporated. This paper presents a distributed implementation of an integer programming approach for solving large-scale ATFM problems in the presence of unmanned aircraft. Given desired mission plans and flight-specific operating and delay costs, the proposed approach uses column generation to determine optimal trajectories in space and time, in the presence of network and flight connectivity constraints, airport and airspace capacity constraints, and geofencing constraints. Using projected demand for the year 2030 from the United States with approximately 48, 000 passenger flights and 29, 000 UAS operations (on a wide range of missions) per day, we show that our implementation can find nearly-optimal trajectories for a 24-hour period in less than 4 minutes. Furthermore, a rolling horizon implementation (with 6-8 hour time windows) results in run times of less than a minute. In addition to being the largest instances of the ATFM problem solved to date, these results represent the first effort to incorporate UAS trajectories into airspace and airport resource sharing problems.