Algorithms For Recovery Of Undefined Input Effects Of Dynamic Systems Under Conditions Of Correlated Noise Of Object
Abstract
The article discusses the issues of formation of stable algorithms for estimation of uncertain
input effects of dynamic systems under conditions of correlated noise of the object. Principle of
expansion of mathematical models is used for formation of initial equation taking into account
correlation of object noise. A two-step computing procedure is used to calculate the Kalman
filter gain. To solve the original equation, the methods of Tikhonov, singular decomposition of
the matrix and Lavrentiev are used. Provides expressions for selecting the regularization option.
These algorithms allow to regulate the task of restoring undefined input effects in dynamic
systems under conditions of correlated noise of the object.