Worst-Case Maximum Likelyhood Estimation in the Linear Model

This paper addresses the problem of maximum likelyhood parameter estimation in linear models aaected by structured deterministic uncertainty in the regression matrix and random gaussian noise. The proposed estimator maximizes the worst-case (with respect to the deterministic uncertainty) likelyhood of the measured sample. The estimate is computed solving a semideenite optimization problem (SDP). In the particular case of unstructured uncertainty, this SDP simply requires minimization of a scalar convex function of one variable.

[1]  A. N. Tikhonov,et al.  Solutions of ill-posed problems , 1977 .

[2]  L. Eldén Perturbation Theory for the Least Squares Problem with Linear Equality Constraints , 1980 .

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

[4]  A. Tits,et al.  Robustness in the presence of mixed parametric uncertainty and unmodeled dynamics , 1991 .

[5]  Å. Björck Component-wise perturbation analysis and error bounds for linear least squares solutions , 1991 .

[6]  J. Bunch,et al.  Collinearity and Total Least Squares , 1994, SIAM J. Matrix Anal. Appl..

[7]  K. Poolla,et al.  Identification of linear parameter-varying systems via LFTs , 1996, Proceedings of 35th IEEE Conference on Decision and Control.

[8]  Ali H. Sayed,et al.  A fast iterative solution for worst-case parameter estimation with bounded model uncertainties , 1997, Proceedings of the 1997 American Control Conference (Cat. No.97CH36041).

[9]  Minyue Fu,et al.  The real structured singular value is hardly approximable , 1997, IEEE Trans. Autom. Control..

[10]  Laurent El Ghaoui,et al.  Robust Solutions to Least-Squares Problems with Uncertain Data , 1997, SIAM J. Matrix Anal. Appl..

[11]  L. El Ghaoui,et al.  Measurement-scheduled control for the RTAC problem , 1997, Proceedings of the 1997 American Control Conference (Cat. No.97CH36041).

[12]  E. Yaz Linear Matrix Inequalities In System And Control Theory , 1998, Proceedings of the IEEE.

[13]  G. Golub,et al.  Parameter Estimation in the Presence of Bounded Data Uncertainties , 1998 .