Some insights on the choice of the future horizon in closed-loop CCA-type Subspace Algorithms

The length of the future horizon is one of the principal "user choices" in subspace identification. It is well known that for the CCA algorithm the asymptotic variance decreases as the future horizon increases when input signals are white (or absent). There are also examples, when the input is not white, in which the asymptotic variance is minimized when the future horizon is equal to the system order. In this paper we shall discuss how (and why) the choice of the future horizon influences the accuracy. We believe the considerations contained in this paper can be a starting point towards a completely automated choice of user parameters in subspace identification. As an intermediate result we show the a version of the CCA algorithm introduced by Larimore is asymptotically equivalent to a number of recently studied methods, complementing recent results appeared in the literature. The setup we consider includes closed loop identification.

[1]  Lennart Ljung,et al.  Closed-Loop Subspace Identification with Innovation Estimation , 2003 .

[2]  S. Joe Qin,et al.  A NEW SUBSPACE IDENTIFICATION METHOD FOR CLOSED-LOOP SYSTEMS , 2006 .

[3]  T. Ferguson.  A Course in Large Sample Theory , 1996 .

[4]  Alessandro Chiuso.  ASYMPTOTIC EQUIVALENCE OF CERTAIN CLOSED LOOP SUBSPACE IDENTIFICATION METHODS , 2006 .

[5]  H. Akaike.  Markovian Representation of Stochastic Processes by Canonical Variables , 1975 .

[6]  Lennart Ljung,et al.  Some facts about the choice of the weighting matrices in Larimore type of subspace algorithms , 2002, Autom..

[7]  Alessandro Chiuso,et al.  The role of vector autoregressive modeling in predictor-based subspace identification , 2007, Autom..

[8]  C. W. J. Granger,et al.  Economic Processes Involving Feedback , 1963, Inf. Control..

[9]  H. Hotelling.  Relations Between Two Sets of Variates , 1936 .

[10]  Alessandro Chiuso,et al.  On the ill-conditioning of subspace identification with inputs , 2004, Autom..

[11]  Bart De Moor,et al.  N4SID: Subspace algorithms for the identification of combined deterministic-stochastic systems , 1994, Autom..

[12]  Alessandro Chiuso,et al.  Consistency analysis of some closed-loop subspace identification methods , 2005, Autom..

[13]  Dietmar Bauer,et al.  Comparing the CCA Subspace Method to Pseudo Maximum Likelihood Methods in the case of No Exogenous Inputs , 2005 .

[14]  Alessandro Chiuso,et al.  Constructing the state of random processes with feedback , 2003 .

[15]  Lennart Ljung,et al.  ON THE ROLE OF FUTURE HORIZON IN CLOSED-LOOP SUBSPACE IDENTIFICATION , 2006 .

[16]  Lennart Ljung,et al.  Subspace identification from closed loop data , 1996, Signal Process..

[17]  John F. MacGregor,et al.  A framework for subspace identification methods , 2001, Proceedings of the 2001 American Control Conference. (Cat. No.01CH37148).

[18]  Magnus Jansson,et al.  Subspace Identification and ARX Modeling , 2003 .

[19]  Wallace E. Larimore,et al.  Large Sample Efficiency for Adaptx Subspace System Identification with Unknown Feedback , 2004 .

[20]  W. Larimore.  System Identification, Reduced-Order Filtering and Modeling via Canonical Variate Analysis , 1983, 1983 American Control Conference.

[21]  Michel Verhaegen,et al.  Identification of the deterministic part of MIMO state space models given in innovations form from input-output data , 1994, Autom..

[22]  Dietmar Bauer,et al.  Asymptotic properties of subspace estimators , 2005, Autom..

[23]  G. Reinsel,et al.  Prediction of multivariate time series by autoregressive model fitting , 1985 .

[24]  A. Chiuso.  The Role of Vector AutoRegressive Modeling in Subspace Identification , 2006, Proceedings of the 45th IEEE Conference on Decision and Control.

[25]  Alessandro Chiuso,et al.  Numerical conditioning and asymptotic variance of subspace estimates , 2004, Autom..

[26]  B. Moor,et al.  Subspace identification for linear systems , 1996 .

[27]  Alessandro Chiuso,et al.  On the relation between CCA and predictor-based subspace identification. , 2005, CDC 2005.

[28]  A. Chiuso,et al.  PREDICTION ERROR VS SUBSPACE METHODS IN CLOSED LOOP IDENTIFICATION , 2005 .

[29]  Wallace E. Larimore,et al.  Canonical variate analysis in identification, filtering, and adaptive control , 1990, 29th IEEE Conference on Decision and Control.

[30]  Bart De Moor,et al.  Subspace algorithms for the stochastic identification problem, , 1993, Autom..

[31]  Alessandro Chiuso.  Asymptotic variance of closed-loop subspace identification methods , 2006, IEEE Transactions on Automatic Control.