Title :
Asymptotic variance of a certain closed-loop subspace identification method
Author :
Chiuso, Alessandro
Author_Institution :
Dept. of Inf. Eng., Padova Univ., Italy
Abstract :
In this paper we study the asymptotic properties of the "innovation estimation" algorithm by Qin and Ljung. We derive expressions for the asymptotic error and decompose it in a "bias" term plus a "variance" term. The asymptotic variance formulas are rather complicated in the general case. We report expression under a simplifying assumption.
Keywords :
closed loop systems; errors; parameter estimation; asymptotic error; asymptotic properties; asymptotic variance formula; bias term; closed-loop subspace identification method; innovation estimation algorithm; variance term; Algorithm design and analysis; Feedback loop; Recursive estimation; Signal processing; State estimation; Stochastic processes; Stochastic systems; Stress control; System identification; Technological innovation;
Conference_Titel :
Decision and Control, 2004. CDC. 43rd IEEE Conference on
Conference_Location :
Nassau
Print_ISBN :
0-7803-8682-5
DOI :
10.1109/CDC.2004.1430375
