The Regulation of Restoration Algorithms

Svetlana Ivanova and Robert Ashinyants

We consider the problem of entrance signal restoration if we know the exit
signal in the case of the bad conditionality of the correlation noise matrix.
We offer to solve this problem using the regularization method. Reverse
problems concern a wide class of incorrect raised problems. Problem of a
entrance signal restoration in measuring of the exit signal is incorrect
raised too. If correlation noise matrix is degenerated, it's impossible to
use classic method of Kalman filter. In the general case this method can't
use for definition of restoration system parameters unambiguously and
guaranty its stability. But often this method gives very effective solutions.
We also demonstrate applicatibility computer algebra to a simple control
system at the presense of noise and computer the results obtained with and
without computer algebra.