Neuromorphic estimation of the parameters of the object’s motion according to the positioning data of GLONASS

Abstract

The article is devoted to the problem of expanding the capabilities of onboard GLONASS. In GLONASS, the implementation of various methods is possible to determine the parameters of the object’s trajectory. Pseudo-range method is supplemented with well-known error compensation methodologies. In particular, a two-frequency error determination method introduced to compensate errors of radio signal passed through the ionosphere. This makes it possible to solve the problem of precise estimation of object’s location coordinates very effectively. It is actual to consider the location coordinates as the initial information in the construction of onboard navigation algorithms for estimating other parameters of the trajectory, among which the most important is the velocity vector of the object relative to the Earth surface. The article presents a mathematical model of the inverse trajectory problem, the purpose of which is to evaluate object’s location coordinates derivatives, described the used technology, research is carried out and procedures are proposed to improve the solvability of the problem under conditions of finite accuracy of measurements and representation of numbers in a computing environment. To solve the problem, a neural-like algorithm of the Kalman type is proposed. The results of computational experiments are also presented.