Gradient-based identification methods for Hammerstein nonlinear ARMAX models. (English) Zbl 1134.93321

Summary: Two identification algorithms, an iterative gradient and a recursive stochastic gradient based, are developed for a Hammerstein nonlinear ARMAX model, a linear dynamical block following a memoryless nonlinear block. The basic idea is to use the gradient search principle, to replace unmeasurable noise terms in the information vectors by their estimates, and to compute iteratively or recursively the noise estimates based on the obtained parameter estimates. Convergence analysis of the recursive stochastic gradient algorithm indicates that the parameter estimation error consistently converges to zero under certain conditions. The simulation results show the effectiveness of the proposed algorithms.


93B30 System identification
93E03 Stochastic systems in control theory (general)
93C23 Control/observation systems governed by functional-differential equations
93E25 Computational methods in stochastic control (MSC2010)
Full Text: DOI


[1] Narendra, K. S. and Gallman, P. G., ’An iterative method for the identification of nonlinear systems using a Hammerstein model’, IEEE Transactions on Automatic Control 11(3), 1966, 546–550. · doi:10.1109/TAC.1966.1098387
[2] Stoica, P., ’On the convergence of an iterative algorithm used for Hammerstein system identification’, IEEE Transactions on Automatic Control 26(4), 1981, 967–969. · doi:10.1109/TAC.1981.1102761
[3] Rangan, S., Wolodkin, G., and Poolla, K., ’Identification methods for Hammerstein systems’, in Proceedings of the 34th IEEE Conference on Decision and Control, New Orleans, LA, 1995, pp. 697–702.
[4] Haist, N. D., Chang, F., and Luus, R., ’Nonlinear identification in the presence of correlated noise using a Hammerstein model’, IEEE Transactions on Automatic Control 18(5), 1973, 553–555. · doi:10.1109/TAC.1973.1100402
[5] Cerone, V. and Regruto, D., ’Parameter bounds for discrete-time Hammerstein models with bounded output errors’, IEEE Transactions on Automatic Control 48(10), 2003, 1855–1860. · Zbl 1364.93417 · doi:10.1109/TAC.2003.817945
[6] Bai, E. W., ’An optimal two-stage identification algorithm for Hammerstein–Wiener nonlinear systems’, Automatica 34(3), 1998, 333–338. · Zbl 0915.93018 · doi:10.1016/S0005-1098(97)00198-2
[7] Bai, E. W., ’Identification of linear systems with hard input nonlinearities of known structure’, Automatica 38(5), 2002, 853–860. · Zbl 1010.93032 · doi:10.1016/S0005-1098(01)00281-3
[8] Bai, E. W., ’A blind approach to the Hammerstein–Wiener model identification’, Automatica 38(6), 2002, 967–979. · Zbl 1012.93018 · doi:10.1016/S0005-1098(01)00292-8
[9] Wigren, T. and Nordsjö, A. E., ’Compensation of the RLS algorithm for output nonlinearities’, IEEE Transactions on Automatic Control 44(10), 1999, 1913–1918. · Zbl 0956.93068 · doi:10.1109/9.793736
[10] Chang, F. and Luus, R., ’A noniterative method for identification using Hammerstein model’, IEEE Transactions on Automatic Control 16(5), 1971, 464–468. · doi:10.1109/TAC.1971.1099787
[11] Nešić, D. and Mareels, I. M. Y., ’Dead-beat control of simple Hammerstein models’, IEEE Transactions on Automatic Control 43(8), 1998, 1184–1188. · Zbl 0957.93036 · doi:10.1109/9.704998
[12] Ding, F. and Chen, T., ’Identification of Hammerstein nonlinear ARMAX systems’, Automatica 41(9), 2005, 1479–1489. · Zbl 1086.93063 · doi:10.1016/j.automatica.2005.03.026
[13] Bai, E. W., ’Decoupling the linear and nonlinear parts in Hammerstein model identification’, Automatica 40(4), 2004, 671–676. · Zbl 1168.93328 · doi:10.1016/j.automatica.2003.11.007
[14] Pawlak, M., ’On the series expansion approach to the identification of Hammerstein system’, IEEE Transactions on Automatic Control 36(6), 1991, 763–767. · doi:10.1109/9.86954
[15] Ninness, B. and Gibson, S., ’Quantifying the accuracy of Hammerstein model estimation’, Automatica 38(12), 2002, 2037–2051. · Zbl 1023.93067 · doi:10.1016/S0005-1098(02)00101-2
[16] Vörös, J., ’Recursive identification of Hammerstein systems with discontinuous nonlinearities containing dead-zones’, IEEE Transactions on Automatic Control 48(12), 2003, 2203–2206. · Zbl 1364.93172 · doi:10.1109/TAC.2003.820146
[17] Gallman, P. G., ’A comparison of two Hammerstein model identification algorithms’, IEEE Transactions on Automatic Control 21(1), 1976, 124–126. · Zbl 0317.93067 · doi:10.1109/TAC.1976.1101123
[18] Ljung, L., System Identification: Theory for the User, 2nd edn., Prentice-Hall, Englewood Cliffs, NJ, 1999. · Zbl 0949.93509
[19] Goodwin, G. C. and Sin, K. S., Adaptive Filtering, Prediction and Control, Prentice-Hall, Englewood Cliffs, NJ, 1984. · Zbl 0653.93001
[20] Ding, F. and Chen, T., ’Combined parameter and output estimation of dual-rate systems using an auxiliary model’, Automatica 40(10), 2004, 1739–1748. · Zbl 1162.93376 · doi:10.1016/j.automatica.2004.05.001
[21] Ding, F. and Chen, T., ’Hierarchical gradient-based identification of multivariable discrete-time systems’, Automatica 41(2), 2005, 315–325. · Zbl 1086.93063 · doi:10.1016/j.automatica.2005.03.026
[22] Ding, F., Shi, Y., and Chen, T., ’Performance analysis of estimation algorithms of non-stationary ARMA processes’, IEEE Transactions on Signal Processing, in press. · Zbl 1373.94569
[23] Lai, T. L. and Wei, C. Z., ’Extended least squares and their applications to adaptive control and prediction in linear systems’, IEEE Transactions on Automatic Control 31(10), 1986, 898–906. · Zbl 0603.93060 · doi:10.1109/TAC.1986.1104138
[24] Guo, L. and Chen, H. F., ’The {Å}ström–Wittenmark self-tuning regulator revisited and ELS-based adaptive trackers’, IEEE Transactions on Automatic Control 36(7), 1991, 802–812. · Zbl 0752.93074 · doi:10.1109/9.85060
[25] Ren, W. and Kumar, P. K., ’Stochastic adaptive prediction and model reference control’, IEEE Transactions on Automatic Control 39(10), 1994, 2047–2060. · Zbl 0827.93071 · doi:10.1109/9.328821
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.