Trajectory prediction via a feature vector approach
| DWPI Title: Method for predicting destination of aircraft, involves identifying destination defined for feature vector, and presenting identified destination defined for feature vector as predicted destination of current trajectory |
| Abstract: Described herein are various technologies pertaining to extracting one or more features from trajectory data recorded during motion of a body, and further, generating a n-dimensional feature vector based upon the one or more extracted features. The n-dimensional feature vector enables expedited analysis of the trajectory data from which the feature vector was generated. For example, rather than having to analyze a trajectory curve comprising a large number of time-position data points, the n-dimensional feature vector can be compared with one or more search parameters to facilitate clustering of the trajectory data associated with the n-dimensional feature vector with other trajectory data which also satisfies the search request. The trajectory data can be plotted on a screen in combination with the n-dimensional feature vector, and other pertinent information. The trajectory data, etc., can be displayed using heat maps or other graphical representation. |
| Use: Method for predicting destination of aircraft. Can also be used in watercraft, vehicle, person, animal, ball, projectile, and object. |
| Advantage: The destination for the aircraft in motion can be realized while the aircraft is still en route. The ability to expeditiously and efficiently process many potential historical matching trajectories facilitates a quantitative approach for describing the probabilities of many potential destinations in a meaningful manner. |
| Novelty: The method involves generating a current feature vector that is representative of a first segment based upon multiple known positions of an aircraft. Multiple historic feature vectors that respectively pertain to multiple historic trajectories that originated are obtained from the first location. The historic feature vector in the historic feature vectors is comprised of a different destination. The current feature vector is compared with the each historic feature vector in the multiple historic feature vectors. One feature vector in the historic feature vectors is determined to have a trajectory (110) that best matches the current trajectory, based upon the comparison. A destination defined for the feature vector is identified. The identified destination defined for the feature vector is presented as a predicted destination of the current trajectory. |
| Filed: 3/9/2017 |
| Application Number: US15454812A |
| Tech ID: SD 13665.1 |
| This invention was made with Government support under Contract No. DE-NA0003525 awarded by the United States Department of Energy/National Nuclear Security Administration. The Government has certain rights in the invention. |
| Data from Derwent World Patents Index, provided by Clarivate All rights reserved. Republication or redistribution of Clarivate content, including by framing or similar means, is prohibited without the prior written consent of Clarivate. Clarivate and its logo, as well as all other trademarks used herein are trademarks of their respective owners and used under license. |