Accuracy of Algorithms of Parametric Identification

A technique is developed for error estimation of the class of algorithms for parametric identification, allowing the structure decomposition to a block-shaper of the matrix system and a block-solver of this system. Theoretical-probabilistic characteristics of the block-shaper in the algorithm of generating functions are investigated during the experimental identification of coefficients of the differential equation of moments of the oscillatory model of the longitudinal short-period movement of the aircraft. A hypothesis of the most unfavorable distribution of signs of errors of the experimental data registration is accepted. Two first probabilistic moments of matrix errors, which predetermine the inherited error of the block-solver, are calculated analytically. The regularization of the identification process is recommended by switching of the filtering unit in parallel. The feasibility of separation of the stage of filtering of experimental measurements of input-output processes from the stage of the identification itself of linear stationary dynamical system is discussed.

References

[1] Eykhoff P. System identification. London, Wiley, 1974. 253 p. (Russ. ed.: Eykkhoff P. Osnovy identifikatsii sistem upravleniya. Moscow, Mir Publ., 1975. 193 p.).

[2] Ljung L. System identification: Theory for the user. New York, Prentice Hall, 1987. 520 p. (Russ. ed.: L’yung L. Identifikatsiya sistem. Teoriya dlya pol’zovatelya. Moscow, Nauka Publ., 1991. 432 p.).

[3] Ivanov V.V. Metody vychisleniy na EVM [Methods of computer calculations]. Kiev, Naukova Dumka Publ., 1986. 584 p.

[4] Forsythe G.E., Moler C.B. Compuer solution of linear algebraic systems. Englewood Cliffs, Prentice Hall, 1967. 167 p. (Russ. ed.: Forsayt Dzh., Moler K. Chislennoe reshenie sistem lineynykh algebraicheskikh uravneniy. Moscow, Mir Publ., 1969. 168 p.).

[5] Lebedev A.A., Chernobrovkin L.S. Dinamika poleta [Flight dynamics]. Moscow, Mashinostroenie Publ., 1978. 615 p.

[6] Zhuravlev Yu.V. Identification of linear time-invariant systems by generating functions. In: Kozlov V.I., Protopopov A.S. Issledovanie, proektirovanie i raschet tekhnicheskikh sredstv kibernetiki, radio-sistem i ustanovok letatel’nykh apparatov [The research, design, and calculation of cybernetics hardware, radio and aircraft systems]. Tr. MAI [Trans. Moscow Aviat. Inst.], vol. 348. Moscow, MAI Publ., 1976. pp. 9-16 (in Russ.).

[7] Loeb J.M., Cahen G.M. Extraction, a partir des enregistrements de mesures, des parameters dynamiques d’un systeme. Automatisme, 1963, no. 12, pp. 479-486.

[8] Tokaya K. The use of Hermite functions for systems identification. IEEE Trans., 1968, vol. AC-13, no. 4.

[9] Nikiforov A.F., Uvarov V.B. Spetsial’nye funktsii matematicheskoy fiziki [Special functions of mathematical physics]. Moscow, Nauka Publ., 1984. 344 p.

[10] Zhuravlev Yu.I., Gurevich I.B. Artificial perception and image recognition. Raspoznavanie, Klassifikatsiya, Prognoz [Recognit. Classif. Predict.], 1989, vol. 2, no. 5, pp. 5-73 (in Russ.).

[11] Galanin M.P., Savenkov E.B. Metody chislennogo analiza matematicheskikh modeley [Methods of numerical analysis of mathematical models]. Moscow, MGTU im. N.E. Baumana Publ., 2010. 591 p.

[12] Tikhonov A.N., Arsenin V.Ya. Metody resheniya nekorrektnykh zadach [Methods for solving ill-posed problems]. Moscow, Nauka Publ., 1986. 288 p.

[13] Zhuravlev Yu.V. Synthesis of adaptive tracking identifier by the direct Lyapunov method. Vestn. Mosk. Gos. Tekh. Univ. im. N.E. Baumana, Estestv. Nauki [Herald of the Bauman Moscow State Tech. Univ., Nat. Sci.], 2010, no. 2, pp. 3-15 (in Russ.).

[14] Zhuravlev Yu.V. Separability of filtration and identification problems for linear time-invariant dynamic systems. In.: Kozlov V.I. Voprosy kibernetiki i radiotekhniki [Problems of cybernetics and radio engineering]. Tr. MAI [Trans. Moscow Aviat. Inst.], vol. 426. Moscow, MAI Publ., 1977. pp. 24-28 (in Russ.).