On local stability in the complete Prony problem

A. A. Lomov

Matematičeskie trudy, Tome 27 (2024) no. 1, pp. 96-138

Résumé

In the variational Prony problem of approximating observations $x$ by the sum of exponentials, expressions are obtained for critical points and second derivatives of the implicit function $\theta(x)$ of exponents' dependence on $x$ data. Upper bounds are proposed for the second order increments $\|\Delta_{2}\theta\|$ with a description of the $\Delta x$ area where $\theta(x)$ is approximately linear. As a consequence, the lower bounds for $\|\Delta\theta\|$ are obtained for small perturbations in $x$. A comparison with the upper bounds for $\|\Delta\theta\|$ by Wilkinson's inequality is given.

Bibliographie

[1] de Prony, Baron Gaspard Riche, “Essai éxperimental et analytique: sur les lois de la dilatabilité de fluides élastique et sur celles de la force expansive de la vapeur de l'alkool, à différentes températures”, Journal de l'École Polytechnique, 1:22 (1795), 24–76

[2] Lomov A. A., “Local stability in the problem of identifying coefficients of a linear difference equation”, Vestn. Novosib. Gos. Univ., Ser. Mat. Mekh. Inform., 10:4 (2010), 81–103

[3] Lantsosh C., Practical methods of applied analysis, PML, M., 1961

[4] Householder A. S., On Prony's method of fitting exponential decay curves and multiple-hit survival curves, Oak Ridge National Lab. Report ORNL-455, Oak Ridge, Tennessee, 1950 | MR | Zbl

[5] Berdyshev, V. I.; Petrak, L. V., Approximation of functions, compression of numerical information, applications, Tr. Inst. Mat. Mekh., Ekaterinburg, 1999

[6] Egorhin A. O., Budyanov V. P., “Smoothing of signals and estimation of dynamic parameters in automatic systems using a digital computer”, Avtometriya, 1973, no. 1, 78–82

[7] Egorhin A. O., “Method of least squares and «fast» algorithms in variational identification and filtering problems (VI method)”, Avtometriya, 1988, no. 1, 30–42

[8] Osborne M. R., “A class of nonlinear regression problems”, Data Representation, eds. R. S. Anderssen and M. R. Osborne, University of Queensland Press, St. Lucia, 1970, 94–101 | Zbl

[9] Markovsky I., Van Huffel S., “Overview of Total Least Squares Methods”, Signal Processing, 87 (2007), 2283–2302 | DOI | Zbl

[10] Ломов А. А., “Variational Identification Methods for Linear Dynamic Systems and Local Extremal Problem”, Upravlenie Bol'shimi Sistemami, 2012, no. 39, 53–94

[11] Egorshin A. O., “Counter equations:smoothing, filtration, identification”, Sib. Elektron. Mat. Izv., 17 (2020), 1322–1351 | MR | Zbl

[12] Ломов А. А., “On the consistency of generalized orthoregressive parameter estimates for a linear dynamical system”, Sib. Zhurn. Industr. Mat, XVI:4(56) (2013), 87–93 | Zbl

[13] Markovsky I., Pintelon R., “Identification of linear time-invariant systems from multiple experiments”, IEEE Trans. Signal Process., 63(13) (2015), 3549–3554 | DOI | MR | Zbl

[14] Osborne M. R., Smyth G. K., “A Modified Prony Algorithm for Exponential Function Fitting”, SIAM Journal of Scientific Computing, 16 (1995), 119–138 | DOI | MR | Zbl

[15] Izmailov A. F., Sensitivity in optimization, Fizmatlit, M., 2006

[16] Lomov A. A., Rusinova E. A., “Comparison of the target functions in the Prony's problem of measurement data approximation”, Bulletin of the South Ural State University. Series: Computational Mathematics and Software Engineering, 11:2 (2022), 18–27

[17] Gantmacher F. R., Theory of matrices, Nauka, M., 1988

[18] Lomov, A. A., Fedoseev, A. V., “Comparison of Parameter Identification Methods for Linear Dynamic Systems Under Mixed Noise”, Sib. Zh. Chist. Prikl. Mat., 18:3 (2018), 45–59 | Zbl

[19] Lomov A. A., “Estimation of trends and identification of time series dynamics in short observation sections”, Izv. Ross. Akad. Nauk, Teor. Sist. Upr., 2009, no. 1, 25–37

[20] Wilkinson J. H., Algebraic eigenvalue problem, Nauka, M., 1970