A Model of Competitive Neuromorphic Aggregation of Information in the Integrated Satellite-Inertial Navigation System

The authors of this paper demonstrate how, with the on-board multi-position information received from a satellite navigation system and the inertial information delivered from three dimension accelerometers with orthogonal axes of sensitivity, forming a movable coordinate trihedron, for the latter one the orientation and angular velocity of the rotation in projections on its own axis can be determined. The mathematical model of the inverse problem in the "state-measurement" form is presented as 1) a dynamic group of equations of functioning of the inertial navigation system (without gyroscopes), with the state vector, which includes the coordinates, specific impulses and angular velocities of the instrumental trihedron rotation, and 2) equations for measurement of the object's location coordinates, identified with the coordinates of trihedron tops in the projections on its axis. A dynamic pseudo inversion (solution) of the problem is realized by the neural network, the model of which is based on Kalman-type algorithm in its multi-model presentation, which allows a judgment about a fair competition between the models in the progress of the state vector evaluation. The concept of "nuclear" and "nuclear-free" setting mechanisms of the neural network synaptic coefficients is introduced. Hypothesizes about a possibility of broadcasting of the characteristics, used in the description of an artificial neural network, are introduced in presentations concerning organization and functioning of the populations of the biological ("live") neurons. The results of the computational experiments are presented.

угловая скорость, ньютонометр, навигационная спутниковая система, нейроморфизм, обратная задача, алгоритм решения, angular velocity, accelerometer, satellite navigation system, neuromorphism, inverse problem, algorithm of solution

1. Девятисильный А. С., Числов К. А. Нейросетевая коррекция безгироскопной инерциальной навигационной системы по спутниковой навигационной информации // Мехатроника, автоматизация, управление. 2013. № 11. С. 67-70.

2. Андреев В. Д. Теория инерциальной навигации. Автономные системы. М.: Наука, 1966.

3. Андреев В. Д. Теория инерциальной навигации. Корректируемые системы. М.: Наука, 1967.

4. Калман Р., Фалб П., Арбиб М. Очерки по математической теории систем. М.: Мир, 1971.

5. Девятисильный А. С., Числов К. А. Модель векторной гравиметрии на базе модели интеграции двухкомпонентной инерциальной, спутниковой и астронавигационной систем // Мехатроника, автоматизация, управление. 2012. № 8. С. 61-63.

6. Лескова Н. Л. В стране лилипутов // В мире науки / Scientific American. 2014. № 6. С. 48-53 (Русская версия SA).