N. A. Kuznetsov, A. V. Lubkov, A. I. Yashin, “On consistency of в aye si an parameter estimates in a circuit for adaptive kalman filtering”, Avtomat. i Telemekh., 1981, Issue 4,Pages <nobr>47

Stochastic Systems

On consistency of в aye si an parameter estimates in a circuit for adaptive kalman filtering

N. A. Kuznetsov, A. V. Lubkov, A. I. Yashin

Moscow

Abstract: The paper is concerned with studies of the properties of Bayesian parameter esti mates for linear systems whose characteristics are observed against the background of random noise. With certain values of the parameters unobservable coordinates are be lieved to be estimatable with the use of a Kalman filter. In the case of an unknown parameter which takes on a numerical set of values an adaptive filter of system parameters and coordinates can be used. The necessary and sufficient conditions are formulated for consistency of parameter estimates in terms of coefficients of the ini tial system and characteristics of the approapriate Kalman filters. An example is given.

UDC: 621.391.172

Received: 09.04.1980