Due to the risks associated with unmanned flight operations, unmanned aircraft systems (UASs) are heavily restricted from operating in the vast U.S. National Airspace System (NAS), and face similar restrictions in other countries. They also present hazards in war zones and, while these hazards may be tolerated to some extent, minimizing them is an important goal for the military. Current restrictions significantly constrain UAS applications and missions, limit training opportunities, and increase development time and cost for new UAS platforms and components. One of the key technologies needed to safely integrate UASs and manned vehicles into shared airspace, including the U.S. NAS, is automated Sense and Avoid (SAA). The effectiveness of SAA algorithms can be significantly enhanced with accurate predictions of air vehicle intent. This is particularly true in the terminal area of operations, which is highly structured and therefore conducive to precise predictions of behavior over a relatively long time horizon. This paper describes an architecture for predicting aircraft intent in the terminal area, and demonstrates its effectiveness using flight data.
展开▼
