IDENTIFICATION OF LINEAR SYSTEMS WITH TIME DELAY OPERATING IN A CLOSED LOOP IN THE PRESENCE OF NOISE

摘要

The subject of this paper is on-line identification of a linear dynamical system with a time delay operating in a closed loop in the presence of noise, The method is based on the equation error. The difficulty presented by the time delay is resolved by "delayed state variables". The required parameter estimate is obtained directly from closed loop input-output data by means of "composite state filters". These prevent the increase of the dimension of the parameter vector which otherwise occurs as a result of the loop closure. A "higher order model" is used in the equation error. This serves the purpose of identifying the time delay by means of the excess model parameters. The method is illustrated by simulated examples.