ALGORITHMS FOR SUSTAINABLE RECOVERY OF INPUT INFLUENCE ON THE BASIS OF DYNAMIC FILTRATION METHODS

The problems of constructing algorithms for the sustainable recovery of unknown signals in dynamic control systems are considered. Algorithms for recovering input actions based on dynamic filtering methods and solving incorrectly posed problems are presented. The regularized Cholesky factorization method for symmetric matrices is used as a regular procedure. The above algorithms make it possible to stabilize the matrix inversion procedure when assessing the state of stochastic objects and thereby improve the accuracy of determining the true state vector estimate under perturbation of the object and observer parameters.