Construction of Phase Portraits of Aircraft Angular Motion by Digital Processing of Optical Images

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Proposed method comprises plotting phase portraits of aircraft angular motion by digital processing of its optical image sequence. As an application, the problem of recognition by the formed phase portraits of the model of angular motion of the aircraft is considered. The decision on the aircraft model belonging to a certain class is made by the criterion of conjugacy of the vector representing the observed object with the reference subspace of the class. Vectors are formed by scanning images of phase portraits built from calculated values of angle and angular velocity of aircraft at intervals of aircraft proper motion. Current readings of angular position are calculated from coordinates of projections of given marker points of aircraft in the image, and angular velocity is estimated using the formula of the first central difference. Reference subspaces of classes are formed b y marked vectors representing the corresponding class. Calculated processes of aircraft angle and angular velocity used to form reference vectors are subjected to preliminary processing by Kalman-Bucy filter. An algorithm for calculating coordinates of projections of marker points on images has been developed, which makes it possible to form reference subspaces by modeling. An example of construction of phase portraits and recognition of aircraft angular motion models in pitch channel with different values of state equation parameters is given