Nationwide evaluation of a conflict resolution algorithm

Abstract

A primary function of today's air traffic controllers is monitoring aircraft and resolving predicted conflicts (i.e. losses of minimum separations between aircraft.) Research is under way to develop automated separation assurance technology to enable safe accommodation of significantly higher traffic demands than can be handled by human controllers. This paper presents simulation results for one such automation concept, the Advanced Airspace Concept, for a full day of flight operations over the continental United States. Comparisons of conflict metrics to previous system-wide studies, which did not study resolutions, are presented. Then the regional variations in conflict metrics, such as the number of conflicts, phases of flight and the encounter angles for conflicting flights are presented. Metrics of resolution performance, such as delay incurred, and the number of trial plans required before a successful resolution maneuver is found are also presented. The results show that there are extremely large variations in the range of the resolution metrics, but statistical analysis is provided to show that resolution performance does vary due to conflict characteristics and the number of aircraft in close proximity to conflict locations. In particular, conflicts with one or both flights within 20 minutes of their destination are shown to incur 5.7 times more delay to resolve than other conflicts and higher local flight densities are shown to increase average delay for resolutions. Finally, the resolution algorithm was found to be fully capable of handling all conflicts under the current-day traffic density used in the simulation.