Sustainable Algorithms For Generalized Estimation Of Dynamic Control Objects

The problems of separate and generalized assessment of the state and parameters of controlled objects under conditions of varying degrees of a priori uncertainty are considered. Stable algorithms for the generalized estimation of the state and parameters based on the quasilinearization method and the formal model of motion under conditions of statistical uncertainty, as well as the correction of the results of the generalized estimation, are presented. It is shown that the considered regularized generalized coordinate and parametric estimation algorithms make it possible to recover with sufficient accuracy the extended state vector of a dynamical system. The above algorithms make it possible to stabilize the procedure for estimating the state of stochastic objects and thereby increase the accuracy of determining the true estimate of the state vector when the parameters of the object and observer are perturbed