|
Vol: 54(68) No: 2 / June 2009 Identification of Dynamic Errors-In-Variables Bilinear System Models via Separable Nonlinear Least Squares T. Larkowski Control Theory and Applications Centre, Coventry University, Faculty of Engineering and Computing, CV1 5FB, Coventry, UK, phone: (+44) 24-7688-8972, e-mail: larkowst@coventry.ac.uk, web: http://www.coventry.ac.uk/researchnet/d/502 J. G. Linden Control Theory and Applications Centre, Coventry University, Faculty of Engineering and Computing, CV1 5FB, Coventry, UK K. J. Burnham Control Theory and Applications Centre, Coventry University, Faculty of Engineering and Computing, CV1 5FB, Coventry, UK Keywords: bilinear systems, errors-in-variables, parameter estimation, system identification Abstract An approach for the identification of dynamic single-input single-output time-invariant bilinear errors-in-variables models is proposed. The method is constructed within the framework of the extended bias compensated least squares technique and utilises the separable nonlinear least squares principle to estimate the system parameters together with the variances of the input and output noise. A comprehensive numerical Monte-Carlo simulation study demonstrates the appropriateness and the relatively high noise robustness of the two realisations of the proposed method. References [1] L. Ljung, System Identification - Theory for the User, 2nd Ed. Prentice Hall PTR, New Jersey, USA, 1999. [2] T. Soderstrom, “Errors-in-variables methods in system identification,” in Automatica, vol. 43, no. 6, pp. 939 – 958, 2007. [3] R. R. Mohler and W. J. Kolodziej, “An overview of bilinear system theory and applications,” in IEEE Trans. on Systems, Man, and Cybernetics, vol. 10, no. 10, pp. 683 – 688, 1980. [4] R. R. Mohler, Nonlinear Systems: Applications to Bilinear Control, vol. 2, Prentice Hall, Englewood Cliffs, NJ, 1991. [5] R. R. Mohler and A. Y. Khapalov, “Bilinear control and application to flexible a.c. transmission systems,” J. of Optimization Theory and Applications, vol. 105, no. 3, pp. 621 – 637, 2000. [6] W. Favoreel, B. De Moor, and Overschee, “Subspace identification of bilinear systems subject to white inputs,” IEEE Trans. on Automatic Control, vol. 44, no. 6, pp. 1157 – 1165, 1999. [7] R. R. Mohler and A. Ruberti, Theory and Applications of Variable Structure Systems. New York, USA: Academics, 1972. [8] F. Carravetta, A. Germani, and M. Raimondi, “Polynomial filtering of discrete-time stochastic linear systems with multiplicative state noise,” in IEEE Trans. on Automatic Control, vol. 42(8), pp. 1106 – 1126, 1997. [9] W. L. De Koning, “Optimal estimation of linear discrete-time systems with stochastic parameters,” Automatica, vol. 20, no. 1, pp. 113 – 115, 1984. [10] R. K. Pearson, Discrete-Time Dynamic Models. Oxford University Press, New York, USA, 1999. [11] I. Vajk and J. Hetthéssy, “Identification of nonlinear errors-in-variables models,” Automatica, vol. 39, pp. 1099 – 2107, 2003. [12] A. Kukush, I. Markovsky, and S. Van Huffel, “Consistent estimation in the bilinear multivariate errors-in-variables model,” Metrika, vol. 57, no. 3, pp. 253 – 285, 2003. [13] G. Vandersteen, “On the use of compensated total least squares in system identification,” in IEEE Trans. on Automatic Control, vol. 43, pp. 1436 – 1441, 1998. [14] S. Han, J. Kim, and K. Sung, “Extended generalized total least squares method for the identification of bilinear systems,” in IEEE Trans. on Signal Proc., vol. 44(4) , pp. 1015 – 1018, 1996. [15] T. Larkowski, J. G. Linden, B. Vinsonneau, and K. J. Burnham, “Regularized structured total least norm for the identification of bilinear systems in the errors-in-variables framework,” in Proc. of the 3rd Int.Conf. on Systems, Cancun, Mexico, 2008. [16] T. Larkowski, B. Vinsonneau, and K. J. Burnham, “Bilinear model identification in the errors-in-variables framework via the bias- compensating least squares,” in Proc. CD-ROM IAR and ACD Int. Workshop, Grenoble, France, 2007. [17] T. Larkowski, J. G. Linden, B. Vinsonneau, and K. J. Burnham. Frisch scheme identification for dynamic diagonal bilinear models. (in press) Int. J. of Control, 2009. [18] M. Ekman, “Modeling and control of bilinear systems: Applications to the activated sludge process,” Ph.D. Thesis, Uppsala University, Sweden, 2005. [19] G. H. Golub and V. Peryera, “The differentiation of pseudo-inverses and nonlinear least squares problems whose variables separate,” SIAM J. of Numerical Analysis, vol. 10, no. 2, pp. 413 – 432, 1973. [20] U. Kotta, S. Nomm, and A. S. I. Zinober, “On state space realizability of bilinear systems described by higher order difference equations,” in Proc. of 42nd IEEE Conf. on Decision and Control, vol. 6, pp. 5685 – 5690, 2003. [21] R. Diversi, R. Guidorzi, and U. Soverini, “Yule-Walker equations in the Frisch scheme solution of errors-in-variables identification problems,” in Proc. of the 17th Int. Symposium on Mathematical Theory of Networks and Systems, Kyoto, Japan, 2006. [22] S. Thil, M. Gilson, and H. Garnier, “On instrumental variable-based methods for errors-in-variables model identification,” in Proc. of 17th IFAC World Congress, Seoul, Korea, pp. 426 – 431, 2008. [23] M. Ekman, M. Hong, and T. Soderstrom, “A separable nonlinear least-squares approach for identification of linear systems with errors in variables,” in 14th IFAC Symp. on System Identification, Newcastle, Australia, 2006. [24] M. Ekman, “Identification of linear systems with errors in variables using separable nonlinear least squares,” in Proc. of 16th IFAC World Congress, Prague, Czech Republic, 2005. |