Integration of Inertial and Optical Navigation Systems Based on Stochastic Nonlinear Estimation Methods

To date one of the most highly accurate methods of solving the problem of autonomous navigation is the processing of optical information captured during the movement of an object. Existing optical flow processing methods based on the determination of the so-called velocity field allow us to estimate only projections of the linear and angular velocities of an object. This, in its turn, is only a part of the overall task of navigation – estimating the current coordinates of the object and the parameters of its spatial orientation. Due to the limited capabilities of such optical navigation systems, it is proposed to integrate optical navigation systems, the advantage of which is stable autonomous monitoring of linear and angular motion parameters over an arbitrary time interval (moreover, with minimal hardware costs), Inertial navigation systems functionality provides the solution to the problem of autonomous navigation as a whole. Due to the unavoidable interference of various physical nature, which significantly distorts the measurements of the mentioned NS, the synthesis of the considered integrated inertial optical system was carried out using the methods of modern stochastic filtering theory, which by far are the most effective in assessing state parameters in a noisy environment. An extended Kalman filter, modified to take into account the correlation of object and observer noise, was chosen as an algorithm for estimating the vector of object navigation parameters based on measurements of the developed inertial-optical navigation system. The results of a numerical experiment illustrating the effectiveness of the proposed approach are presented.