A computational approach to conflict detection problems for air traffic control

Formulates a conflict detection problem for two aircraft performing general flight maneuvers. The key ingredient is the development of probabilistic models for flight control errors that arise due to wind, navigational uncertainties, flight control errors and intent. It is assumed that each aircraft is performing a trimmed flight maneuver. Assumptions are made so that the relative position errors are characterized by a Chi-square distribution. The relative position errors of the aircraft are described by an ellipsoid that translates, rotates, and changes shape and size. The conflict detection problem is formulated in terms of suitable separation measures of this ellipsoid and an ellipsoid that defines a protected zone. A separation measure is introduced that provides an upper bound for the conflict probability. A simple and numerically efficient algorithm is given so that quantitative measures of the aircraft separation can be efficiently computed.