Flight trajectory prediction method and device based on Koopman neural network

• Main Authors: LU JING, BAI YIDAN, CHEN RAN
• Format: Patent
• Language: Chinese; English
• Published: 03.10.2023
• Subjects: CONTROLLING; PHYSICS; REGULATING; SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES

The invention relates to the field of flight safety, and discloses a flight trajectory prediction method and device based on a Koopman neural network, and the method comprises the steps: obtaining a standard flight trajectory of each flight subject and standard flight trajectory data, and obtaining a feature difference between a standard representation flight trajectory and the standard flight trajectory of each flight subject; acquiring a flight training trajectory data set, and training the trajectory generation model to obtain a flight trajectory generation model; according to the obtained feature difference, identifying to obtain a corresponding flight subject; the method comprises the following steps: acquiring flight data of an aircraft in real time according to a flight data acquisition device, and obtaining a predicted flight path through a flight path prediction model based on a Koopman neural network; if the flight trajectory similarity is not smaller than a set similarity threshold value, it is jud