Science, Technologies, Innovations №2(22) 2022, 52-60 p

 PDF

http://doi.org/10.35668/2520-6524-2022-2-07

Rozora I. V. — D. Sc. in Physics and Mathematics, Taras Shevchenko National University of Kyiv, 4-d Akademika Glushkova Avenue, Kyiv, Ukraine, 02000; +38 (044) 521-35-35; irozora@knu.ua; ORCID: 0000-0002-8733-7559

Melnyk A. O. — Master Student, Taras Shevchenko National University of Kyiv, 4-d Akademika Glushkova Avenue, Kyiv, Ukraine, 02000; +38 (044) 521-35-35; melinik2011@gmail.com; ORCID: 0000-0002-3167-4353

CONSTRUCTION OF GOODNESS-OF-FIT CRITERIA FOR THE TYPE OF IMPULSE RESPONSE FUNCTION

Abstract. The article is devoted to the study of the impulse response function, its estimation and properties, square-Gaussian random variables and processes, the rate of convergence of the unknown impulse response function, testing the hypothesis about the type of impulse response function, building a simulation model. The study showed that the pulse response function is the output signal of the system during signal processing, when the input signal is a short pulse. In a more general form, the impulse response function describes the response or output of the system as a function of time. Also, the impulse response function is considered a property of linear displacement systems. During the study of the estimation of the impulse response function on orthonormal and trigonometric bases, two conditions A, B and remarks to them were formed, which are used in the future to find different coefficients. The study of square-Gaussian random variables and processes has shown the benefits of using them in relation to the impulse response function. A theorem was also presented, which estimated the probability of a large deviation of the square-Gaussian process in the norm of a continuous function. To study the rate of convergence of the unknown impulse response function in the space of continuous functions and in the space L2, a lemma was formed, as well as a theorem that directly showed the rate of convergence of the impulse response function in the space of continuous functions. Zero and alternative hypotheses were formed. The null hypothesis claimed that the impulse response function existed, and the alternative hypothesis suggested the opposite. To test the hypothesis about the form of the impulse response function, a theorem was used by which a criterion was formed. Visual Studio Community 2022 integrated development environment (C ++ programming language) and Wolfram Mathematica computer algebra system for analytical transformations and numerical calculations were used to build the simulation model, which allowed to make mathematical calculations quite accurately.

Keywords: impulse response function, square-Gaussian quantities, estimation of impulse response function.

REFERENCES

  1. Bendat, J. S., & Piersol, A. G. (1980). Engineering Applications of Correlation and Spectral Analysis. New York: Wiley, 472 p.
  2. Schetzen, M. (1980). The Volterra and Wiener Theories of Nonlinear Systems. New York: Wiley, 550 p.
  3.  Akaike, H. (1965). On the statistical estimation of the frequency response function of a system having multiple input. Annals of the Institute of Statistical Mathematics, 17, P. 185-210. https://doi.org/10.1007/BF02868166
  4. Akaike, H. (1966). On the use of non-Gaussian process in the identification of a linear dynamic system. Annals of the Institute of Statistical Mathematics, 18, P. 269-276. https://doi.org/10.1007/BF02869534
  5. Kozacenko, Yu., & Rozora, I. (2016). On cross-correlogram estimators of impulse response function. Theoretical Probability and Mathematical Statistics, 93, P. 79-91. https://doi.org/10.1090/tpms/995
  6. Kozacenko, Yu., & Rozora, I. (2016).   A criterion for testing hypothesis about impulse response function. Statistics, Optimization & Information Computing, 4, P. 214-232. https://doi.org/10.19139/soic.v4i3.222
  7. Rozora, I. (2018). hypothesis testing for the shape of impulse response function. Communications in Statistics – Theory and Methods, 47(6), P. 1459-1474. https://doi.org/10.1080/03610926.2017.1321125
  8. Kozacenko, Yu., Rozora, I., & Turchyn, Ye. (2007). On an expansion of random processes in series. Random Operators and Stochastic Equations, 15 (1), P. 15-34. https://doi.org/10.1515/ROSE.2007.002
  9. Kozacenko, Yu., Rozora, I., & Turchyn, Ye. (2011). Properties of Some Random Series. Communications in Statistics – Theory and Methods, 40 (19-20), P. 3672-3683. https://doi.org/10.1080/03610926.2011.581188
  10.  Kozacenko, Yu, Pogoriliak, O., Rozora, I., & Tegra, A. (2016). Simulation of Stochastic Processes with given Accuracy and Reliability. London: ISTE Press Ltd and Elsevier Ltd, 346 p.
  11.  Kozacenko, Yu., Pashko, A., & Rozora, I. (2007). Simulation of stochastic processes and fields. Kyiv, 232 p.
  12.  Gikhman, I. I., & Skorokhod, A. V. (1996). Introduction to the theory of random processes. New York: Dover Publications Inc, 544 p.
  13.  Buldygin, V. V., & Kozachenko, Yu. V. (2000). Metric characterization of random variables and random processes. Providence: American Mathematical Society, 257 p. https://doi.org/10.1090/mmono/188
  14.  Kozacenko, Yu., & Moklyachuk,  O. (1999). Large deviation probabilities for square-Gaussian stochastic processes. Extremes, 2 (3), P. 269-293.
  15.  Kozacenko, Yu., & Moklyachuk,  O. (2000). Square-Gaussian stochastic processes. Theory of Stochastic Processes, 3-4, P. 98-121.
  16.  Kozacenko, Yu., & Stus, O. (1998). Square-Gaussian random processes and estimators of covariance functions. Mathematical Communications, 3 (1), P. 83-94.
  17.  Gondenshluger, A., &  Pereverzev,  S. (2003). On adaptive inverse estimation of linear functionals in Hilbert scales. Bernoulli, 9 (5), P. 783-807. https://doi.org/10.3150/bj/1066418878
  18.  Meister, A. (2009). Deconvolution Problems in Nonparametric Statistics. Berlin: Springer-Verlag, 210 p.
  19.  Ingster, Y.,  & Suslina,  I. (2003).  Nonparametric Goodness-of-Fit Testing Under Gaussian Models. Springer, 2,  P. 141-152. https://doi.org/10.1007/978-0-387-21580-8
  20.  Tsybakov, A. (2009). Introduction to Nonparametric Estimation. Springer Series in Statistics. New York: Springer, 214 p. https://doi.org/10.1007/b13794
  21.  Fan, J.,  & Gijbels,  I. (1996). Local Polynomial Modelling and its Applications. London: Chapman & Hall, 360 p.
  22.  Hannan, E., & Deistler,  M. (1988). The Statistical Theory of Linear Systems. Wiley, 8, P. 135-143.
  23.  Cardot, H., & Johannes,  J. (2010).  Thresholding projection estimators in functional linear models. Journal of Multivariate analysis, 101, P. 395-408. https://doi.org/10.1016/j.jmva.2009.03.001
  24.  Pillonetto, G.,  Dinuzzo, F., & Chen,  T. (2014). Kernel methods in system identification, machine learning and function estimation: A survey. Automatica, 40 (3), P. 657-682. https://doi.org/10.1016/j.automatica.2014.01.001
  25.  Delaigle, A.,  Hall, P. & Meister,  A. (2008). On deconvolution with repeated measurements. Annals of Statistics, 36, P. 665-685. https://doi.org/10.1214/009053607000000884
  26.  Dashti, M., & Stuart, A. (2007). The Bayesian approach to inverse problems. New York: Springer, 107 p.
  27.  Alquier, P.,  Gautier, E., & Stoltz, G. (2011). Inverse Problems and High-Dimensional Estimation. New York: Springer, 363 p.