Artabaev Y. A technique for increasing the accuracy of integrated navigation systems based on predictive models using genetic algorithms.

The thesis is devoted to improving the accuracy of navigation definitions of integrated navigation systems, and on this basis - improvement of existing methods, including the development of an own method for correcting an autonomous inertial navigation system (INS) in the original signal based on signals from precession torque sensors, which makes it possible to take into account the nonlinear components of dynamic errors and assessment of its effectiveness. Mathematical models of errors in navigation systems of an aircraft (AC) have been investigated. A systematic analysis of the models is carried out and on its basis the most promising ones for use in correction algorithms are selected. A block of two-stage differentiating gyroscopes is investigated and a simple scheme for correcting dynamic errors is developed, taking into account errors that depend on the cross angular velocities arising during complex motion of the aircraft. The given measurements for the GA are formed, based on a nonlinear model of the ANN errors, used to compensate for the errors of an autonomous INS, predict the route and exit to a given point of the aircraft for object recognition and reconnaissance of the situation. The results of mathematical modeling showed the efficiency of the developed methods. The analysis of the results of mathematical modeling showed that the developed methods make it possible to increase the efficiency of navigation systems in an autonomous mode and to ensure the fulfillment of the assigned tasks of the aircraft.