For the case of unknown parameters that are subject to noise, a successive algorithm for identification of the discrete-time linear stochastic system was proposed. Special selection of a random instant of observation abortion in the least-squares method enabled determination of the theoretical boundary of the root-mean-square precision of the vector estimator which is inversely proportional to the procedure parameter defining its duration. The results of numerical modeling were presented.
ASJC Scopus subject areas
- Control and Systems Engineering