PROJECTIONAL RESEARCH METHODS OF THE REVERSE PROBLEM OF LINEAR DYNAMIC SYSTEMS

Abstract

The possibility of finding the structure of an operator of a dynamic object from its output signals on the basis of the structural properties of linear operators and the ordering of a set of output signals presented in the form of Hankel forms and Hankel matrices is considered. According to the heuristic approach, the input signal acts on an object, while collecting information about all degrees of freedom of a dynamic uncontrolled object. Such an input signal having an infinite spectrum is white noise. The methods for finding the structure of the operator and the estimation of its parameters for the linear case, as well as the method for identifying the model of a multidimensional dynamic system, are considered. Methods have been developed for finding a model of a dynamic system, which is defined only by the output signal, and determining the characteristics of a dynamic system, in contrast to the known processes described by the problems of identification, control and measurement. An illustrated sequence of construction of the model of the linear dynamic system operator as a solution to the inverse dynamics problem: to determine the structure of the operator in the state space based on the output signal, which allows to develop computational algorithms for real dynamical systems in a linear approximation.