Sensitivity of structures for the identification of linear systems from impulse response data

作者: V.E. DeBrunner , A.A. Beex

DOI: 10.1109/ICASSP.1989.266904

关键词: Newton's methodControl theoryDigital filterImpulse responseQuantization (signal processing)Estimation theoryMathematicsRate of convergenceLinear systemTransient response

摘要: The authors investigate the usefulness of structures with high parameter sensitivities for purpose increasing convergence rate identification algorithms. In particular, they examine two SISO (single input, single output) structures: direct form II and dual generalized Hessenberg representation. A Gauss-Newton procedure is presented implementing from impulse response data. then present a sensitivity analysis provide some illustrative examples. >

参考文章(6)
Victor E. Debrunner, A A. (Louis) Beex, Sensitivity analysis of digital filter structures SIAM Journal on Matrix Analysis and Applications. ,vol. 9, pp. 106- 125 ,(1988) , 10.1137/0609009
J. Rissanen, Recursive Identification of Linear Systems Siam Journal on Control. ,vol. 9, pp. 335- 341 ,(1971) , 10.1137/0309031
William B. Gragg, Anders Lindquist, On the partial realization problem Linear Algebra and its Applications. ,vol. 50, pp. 277- 319 ,(1983) , 10.1016/0024-3795(83)90059-9
A. Beex, Efficient generation of ARMA cross convariance sequences international conference on acoustics, speech, and signal processing. ,vol. 10, pp. 327- 330 ,(1985) , 10.1109/ICASSP.1985.1168406
D. K. LINDNER, Near aggregation, the dual GHR and pole-zero cancellation International Journal of Control. ,vol. 48, pp. 705- 727 ,(1988) , 10.1080/00207178808906206
A.A. Beex, V.E. Debrunner, Reduced sensitivities of direct form digital (sub) filter structures by increasing system order IEEE Transactions on Circuits and Systems. ,vol. 36, pp. 438- 442 ,(1989) , 10.1109/31.17592