Combined instrumental variable and subspace fitting approach to parameter estimation of noisy input-output systems

The paper considers the problem of estimating the parameters of linear discrete-time systems from noise-corrupted input-output measurements, under fairly general conditions: the output and input noises may be auto-correlated and they may be cross-correlated as well. By using the instrumental-variable (IV) principle a covariance matrix is obtained, the singular vectors of which bear complete information on the parameters of the system under study. A weighted subspace fitting (WSF) procedure is then employed on the sample singular vectors to derive estimates of the parameters of the system. The combined IV-WSF method proposed in the present paper is noniterative and simple to use. Its large-sample statistical performance is analyzed in detail and the theoretical results so obtained are used to predict the behavior of the method in samples with practical lengths. Several numerical examples are included to show the agreement between the theoretically predicted and the empirically observed performances. >

[1]  Anthony J. Jakeman,et al.  An instrumental variable method for model order identification , 1980, Autom..

[2]  Bjorn Ottersten,et al.  Array processing in correlated noise fields using instrumental variables and subspace fitting , 1992, [1992] Conference Record of the Twenty-Sixth Asilomar Conference on Signals, Systems & Computers.

[3]  Jitendra K. Tugnait,et al.  Stochastic system identification with noisy input using cumulant statistics , 1990, 29th IEEE Conference on Decision and Control.

[4]  J. Cadzow,et al.  Algebraic approach to system identification , 1986, IEEE Trans. Acoust. Speech Signal Process..

[5]  T. Söderström,et al.  Bias correction in least-squares identification , 1982 .

[6]  M. Levin.  Estimation of a system pulse transfer function in the presence of noise , 1964 .

[7]  W. Zheng,et al.  Identification of a class of dynamic errors-in-variables models , 1992 .

[8]  Umberto Soverini,et al.  The frisch scheme in dynamic system identification , 1990, Autom..

[9]  Carlos E. Davila.  Total least squares system identification and frequency estimation for overdetermined model orders , 1993, 1993 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[10]  T. Söderström,et al.  Optimal instrumental variable estimation and approximate implementations , 1983 .

[11]  Peter C. Young,et al.  Recursive Estimation and Time Series Analysis , 1984 .

[12]  Peter E. Wellstead,et al.  An instrumental product moment test for model order estimation , 1978, Autom..

[13]  Sabine Van Huffel,et al.  Total least squares problem - computational aspects and analysis , 1991, Frontiers in applied mathematics.

[14]  K. Fernando,et al.  Identification of linear systems with input and output noise: the Koopmans-Levin method , 1985 .

[15]  M. Aoki,et al.  On a priori error estimates of some identification methods , 1970 .

[16]  T. Söderström,et al.  Instrumental variable methods for system identification , 1983 .

[17]  Petre Stoica,et al.  On the uniqueness of prediction error models for systems with noisy input-output data , 1987, Autom..

[18]  Bjorn Ottersten,et al.  Exact and Large Sample ML Techniques for Parameter Estimation and Detection in Array Processing , 1993 .

[19]  S. S. Wilks,et al.  Linear Regression Analysis of Economic Time Series. , 1938 .

[20]  B. Anderson,et al.  Identifiability in dynamic errors-in-variables models , 1983, The 22nd IEEE Conference on Decision and Control.

[21]  Torsten Söderström,et al.  Identification of stochastic linear systems in presence of input noise , 1981, Autom..