To improve the safety and efficiency of air traffic control, it is important to detect any deviations (or maneuvers) of an aircraft from its planned route. In addition to the detection of maneuvers, the characterization of maneuvers can enhance the situational awareness of air traffic controllers. In this paper, an algorithm for the inference of aircraft maneuvers from flight plans and flight tracks is proposed that identifies when the maneuver starts and ends, what type of maneuver is taken, and how an aircraft takes the maneuver. The aircraft maneuvers in current operations are modeled as a sequence of simple or elementary motions, called intents, and the intents are inferred based on the continuous and discrete states of an aircraft obtained from a hybrid estimation algorithm. The proposed algorithm is demonstrated with flight data from the Airspace Concept Evaluation System, which is a multifidelity modeling and simulation system with full gate-to-gate representation of all the major components of the National Airspace System.
Read full abstract