On Stability of Guaranteed Estimation Problems: Error Bounds for Information Domains and Experimental Design
The two guaranteed estimation problems related to semi-infinite programming theory are considered. For abstract estimation problem in a Banach space we study the dependence of information domains (, ) on measurement errors with intensity of errors tending to zero. The upper estimates for the rate of convergence of information domains to their limit in the Hausdorff metric are given. The experimental design problem for estimation of distributed system with uncertain parameters through available measurements is also considered in the context of guaranteed estimation theory. For the stationary sensor placement problem we describe its reduction to a nonlinear programming problem. In the case of sufficiently large number of sensors it is shown that the solution may be obtained by solving linear semi-infinite programming problem.
KeywordsOptimal Control Problem Uncertain Parameter Uncertain System Linear Continuous Operator Sensor Placement
Unable to display preview. Download preview PDF.
- M.I. Gusev. On the optimality of linear algorithms in guaranteed estimation. In A.B. Kurzhanski and V.M. Veliov, editors, Modeling Techniques for Uncertain Systems, Birkhäuser, Boston, pages 93–110, 1994.Google Scholar
- M.I. Gusev. On Stability of Information Domains in Guaranteed Estimation Problem. In Proceedings of the Steklov Institute of Mathematics, Suppl. 1, MAIK “Nauka/Interperiodika”, pages 104–118, 2000.Google Scholar
- V.K. Ivanov, V.V. Vasin and V.P. Tanana. Theory of Linear Ill-Posed Prolems and Applications, Nauka, Moscow, (in Russian) 1978 .Google Scholar
- A. B. Kurzhanski. Control and Observation under Uncertainty Conditions, Nauka, Moscow, (in Russian) 1977.Google Scholar
- A. B. Kurzhanski and A.Yu. Khapalov. Estimation of distributed fields according to results of observations. In Partial Differential Equations Proceedings of the International Conference in Novosibirsk 1983, pages 102–108, (in Russian) 1986.Google Scholar
- A.B. Kurzhanski and I.F. Sivergina. On noninvertible evolutionary systems: guaranteed estimation and regularization problems, Dokl. Acad. Nauk SSSR, 314:292–296(in Russian) , 1990.Google Scholar
- A.B. Kurzhanski and I.F. Sivergina. Method of guaranteed estimates and regularization problems for evolutionary system, Journal vichislit. rnatem. i mat. phiziki., 32:1720–1733(in Russian) , 1992.Google Scholar
- G.I. Marchuk, Mathematical Modelling and Environmental Problems. Nauka, Moscow, (in Russian) 1982.Google Scholar
- V. Pokotilo. Necessary Conditions for Measurements Optimization. In A.B Kurzhanski and V.M. Veliov, editors, Modeling Techniques for Uncertain Systems, Birkhäuser, Boston, pages 147–159, 1994.Google Scholar
- A.N. Tikhonov and V.Ya. Arsenin. Methods of Solutions of Ill-Posed Problems, Nauka, Moscow, (in Russian) 1974.Google Scholar
- D. Ucinski. Measurement Optimization for Parameter Estimation in Distributed Systems, Technical University Press, Zielona Gora, Poland, 1999.Google Scholar
- D.J. Ucinski, J. Korbicz and M. Zaremba. On optimization of sensors motions in parameter identification of two-dimensional distributed systems, In Proc. 2nd European Control Con ference, pages 1359–1364, Groningen, The Netherlands, 1992.Google Scholar