scientific article; zbMATH DE number 3854942

From MaRDI portal
Publication:3323822





[For the entire collection see Zbl 0527.00026.] This clearly-written paper deals with the identification of closed-loop systems. The basic assumptions are that the system is linear, the controller is also linear and a linear stationary exogeneous process is injected into the loop. The task is to estimate the system dynamics from input-output data. Two estimation methods are considered: the output error method (OEM) and the more complicated prediction error method (PEM). [the author calls them simple least-squares (LS) and weighted least-squares; however in the opinion of this reviewer OEM is a better name for the first method in order to avoid confusion with the ordinary linear LS] A frequency-domain analysis of the consisting properties of the two methods is presented. The main goal is to derive conditions under which OEM and PEM asymptotically lead to the same estimate of the system dynamics.




Cited in
(22)








This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3323822)