×

Robust maximum-likelihood estimation of multivariable dynamic systems. (English) Zbl 1087.93054

Summary: This paper examines the problem of estimating linear time-invariant state-space system models. In particular, it addresses the parametrization and numerical robustness concerns that arise in the multivariable case. These difficulties are well recognised in the literature, resulting (for example) in extensive study of subspace-based techniques, as well as recent interest in ‘data driven’ local co-ordinate approaches to gradient search solutions. The paper here proposes a different strategy that employs the expectation-maximisation (EM) technique. The consequence is an algorithm that is iterative, with associated likelihood values that are locally convergent to stationary points of the (Gaussian) likelihood function. Furthermore, theoretical and empirical evidence presented here establishes additional attractive properties such as numerical robustness, avoidance of difficult parametrization choices, the ability to naturally and easily estimate non-zero initial conditions, and moderate computational cost. Moreover, since the methods here are maximum-likelihood based, they have associated known and asymptotically optimal statistical properties.

MSC:

93E10 Estimation and detection in stochastic control theory
93C35 Multivariable systems, multidimensional control systems
PDFBibTeX XMLCite
Full Text: DOI Link

References:

[1] Bergboer, N. H., Verdult, V., & Verhaegen, M. H. G. (2002). An efficient implementation of maximum likelihood identification of LTI state-space models by local gradient search. In Proceedings of the 41st IEEE CDC; Bergboer, N. H., Verdult, V., & Verhaegen, M. H. G. (2002). An efficient implementation of maximum likelihood identification of LTI state-space models by local gradient search. In Proceedings of the 41st IEEE CDC
[2] Borran, M. J., & Aazhang, B. (2002). EM-based multiuser detection in fast fading multipath environments. In Proceedings of EURASIP; Borran, M. J., & Aazhang, B. (2002). EM-based multiuser detection in fast fading multipath environments. In Proceedings of EURASIP · Zbl 1015.94510
[3] Boyles, R. A., On the convergence of the EM algorithm, Journal of the Royal Statistical Society, Series B, 45, 1, 47-50 (1983) · Zbl 0508.62030
[4] Dempster, A. P.; Laird, N. M.; Rubin, D. B., Maximum likelihood from incomplete data via the EM algorithm, Journal of the Royal Statistical Society, Series B, 39, 1-38 (1977) · Zbl 0364.62022
[5] Dennis, J. E.; Schnabel, R. B., Numerical methods for unconstrained optimization and nonlinear equations (1983), Prentice-Hall: Prentice-Hall Englewood Cliffs, NJ · Zbl 0579.65058
[6] Golub, G.; Loan, C. V., Matrix computations (1989), Johns Hopkins University Press: Johns Hopkins University Press Baltimore, MD · Zbl 0733.65016
[7] Goodwin, G. C.; Feuer, A., Estimation with missing data, Mathematical and Computer Modelling of Dynamical Systems, 5, 3, 220-244 (1999) · Zbl 0940.93069
[8] Isaksson, A., Identification of ARX models subject to missing data, IEEE Transactions on Automatic Control, 38, 5, 813-819 (1993) · Zbl 0785.93028
[9] Jazwinski, A. H., Stochastic processes and filtering theory (1970), Academic Press: Academic Press New York · Zbl 0203.50101
[10] Kailath, T.; Sayed, A. H.; Hassabi, B., Linear estimation (2000), Prentice-Hall: Prentice-Hall Upper Saddle River, NJ
[11] Ljung, L., System identification: Theory for the user (1999), Prentice-Hall, Inc: Prentice-Hall, Inc New Jersey
[12] Ljung, L. (2003). Aspects and experiences of user choices in subspace identification methods. In Proceedings of the 13th IFAC symposium on system identification; Ljung, L. (2003). Aspects and experiences of user choices in subspace identification methods. In Proceedings of the 13th IFAC symposium on system identification
[13] Ljung, L. (2004). MATLAB system identification toolbox users guide, Version 6; Ljung, L. (2004). MATLAB system identification toolbox users guide, Version 6
[14] Magnus, J. R.; Neudecker, H., Matrix differential calculus with applications in statistics and econometrics (1988), Wiley: Wiley Chichester, UK · Zbl 0651.15001
[15] Mathworks (2004). MATLAB users guide, Version 7; Mathworks (2004). MATLAB users guide, Version 7
[16] McKelvey, T.; Helmersson, A.; Ribarits, T., Data driven local co-ord’s for multivariable linear systems with application to system identification, Automatica, 40, 9, 1629-1635 (2004) · Zbl 1055.93019
[17] Meng, X.-L.; Rubin, D. B., On the global and componentwise rates of convergence of the EM algorithm, Linear Algebra and its Applications, 199, 413-425 (1994) · Zbl 0818.65153
[18] Pintelon, R.; Schoukens, J.; McKelvey, T.; Rolain, Y., Minimum variance bounds for overparameterized models, IEEE Transactions on Automatic Control, 41, 5, 719-720 (1996) · Zbl 0850.93863
[19] Rao, C. R., Linear statistical inference and its application (1965), Wiley: Wiley New York · Zbl 0137.36203
[20] Shumway, R. H., An approach to time series smoothing and forecasting using the EM algorithm, Journal of Time Series Analysis, 3, 4, 253-264 (1982) · Zbl 0502.62085
[21] Shumway, R. H.; Stoffer, D. S., Time series analysis and its applications (2000), Springer: Springer Berlin · Zbl 0502.62085
[22] Söderström, T.; Stoica, P., System identification (1989), Prentice-Hall: Prentice-Hall New York · Zbl 0714.93056
[23] Wu, C. F., On the convergence properties of the EM algorithm, The Annals of Statistics, 11, 1, 95-103 (1983) · Zbl 0517.62035
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.