Optimizing trajectory of unmanned aerial vehicles

Abstract

The present disclosure is directed toward methods, non-transitory computer-readable media, and systems for trajectory optimization in a high-altitude, long-endurance aircraft. For example, the systems described herein can generate an optimized flight plan for an aircraft during active flight of the aircraft by utilizing a greedy algorithm with buffering. In one or more embodiments, the systems described herein identify a plurality of possible states and select a predetermined number of the top possible states (based on energy change associated with transitioning to each possible state) at each incremental time period within a flight time window starting from an initial state to a plurality of possible states for a final incremental time period. Furthermore, in some embodiments, the systems described herein select a final state based on a final energy associated with the final state and determine a flight plan for the aircraft from the initial state to the selected final state.