Abstract: A novel separation identification strategy for the neural fuzzy Wiener–Hammerstein system using hybrid signals is developed in this study. The Wiener–Hammerstein system is described by a model consisting of two linear dynamic elements with a nonlinear static element in between. The static nonlinear element is modeled by a neural fuzzy network (NFN) and the two linear dynamic elements are modeled by an autoregressive exogenous (ARX) model and an autoregressive (AR) model, separately. When the system input is Gaussian signals, the correlation technique is used to decouple the identification of the two linear dynamic elements from the nonlinear element. First, based on the input and output of Gaussian signals, the correlation analysis technique is used to identify the input linear element and output linear element, which addresses the problem that the intermediate variable information cannot be measured in the identified Wiener–Hammerstein system. Then, a zero-pole match method is adopted to separate the parameters of the two linear elements. Furthermore, the recursive least-squares technique is used to identify the nonlinear element based on the input and output of random signals, which avoids the impact of output noise. The feasibility of the presented identification technique is demonstrated by an illustrative simulation example and a practical nonlinear process. Simulation results show that the proposed strategy can obtain higher identification precision than existing identification algorithms.

基于混合信号的神经模糊Wiener-Hammerstein系统辨识

[1]Ase H, Katayama T, 2015. A subspace-based identification of Wiener–Hammerstein benchmark model. Contr Eng Pract, 44:126-137.

[2]de Moor B, de Gersem P, de Schutter B, et al., 1997. DAISY: a database for identification of systems. Comput Sci, 38(3):4-5.

[3]dos Santos PL, Ramos JA, de Carvalho JLM, 2012. Identification of a benchmark Wiener–Hammerstein: a bilinear and Hammerstein–Bilinear model approach. Contr Eng Pract, 20(11):1156-1164.

[4]Falck T, Dreesen P, de Brabanter K, et al., 2012. Least-squares support vector machines for the identification of Wiener–Hammerstein systems. Contr Eng Pract, 20(11):1165-1174.

[5]Ghanmi A, Elloumi M, Salhi H, et al., 2020. A recursive hierarchical parametric estimation algorithm for nonlinear systems described by Wiener–Hammerstein models. Asian J Contr, 22(3):1065-1074.

[6]Hafsi S, Laabidi K, Ksouri-Lahmari M, 2012. Identification of Wiener–Hammerstein model with multisegment piecewise-linear characteristic. IEEE Mediterranean Electrotechnical Conf, p.5-10.

[7]Han Y, de Callafon RA, 2012. Identification of Wiener–Hammerstein benchmark model via rank minimization. Contr Eng Pract, 20(11):1149-1155.

[8]Janjanam L, Saha SK, Kar R, et al., 2022. Optimal design of cascaded Wiener–Hammerstein system using a heuristically supervised discrete Kalman filter with application on benchmark problems. Expert Syst Appl, 200:117065.

[9]Jia L, Xiong Q, Li F, 2017. Correlation analysis method based SISO neuro-fuzzy Wiener model. J Process Contr, 58:73-89.

[10]Katayama T, Ase H, 2016. Linear approximation and identification of MIMO Wiener–Hammerstein systems. Automatica, 71:118-124.

[11]Ławryńczuk M, 2016. Nonlinear predictive control of dynamic systems represented by Wiener–Hammerstein models. Nonl Dynam, 86(2):1193-1214.

[12]Li F, Li J, Peng DG, 2017. Identification method of neuro-fuzzy-based Hammerstein model with coloured noise. IET Contr Theory Appl, 11(17):3026-3037.

[13]Li F, Zheng T, He NB, et al., 2022. Data-driven hybrid neural fuzzy network and ARX modeling approach to practical industrial process identification. IEEE/CAA J Autom Sin, 9(9):1702-1705.

[14]Li F, Jia L, Gu Y, 2023a. Identification of nonlinear process described by neural fuzzy Hammerstein–Wiener model using multi-signal processing. Adv Manuf, 11:694-707.

[15]Li F, Zhu XJ, He NB, et al., 2023b. Parameter learning for the nonlinear system described by Hammerstein model with output disturbance. Asian J Contr, 25(2):886-898.

[16]Li LW, Ren XM, 2018. Identification of nonlinear Wiener–Hammerstein systems by a novel adaptive algorithm based on cost function framework. ISA Trans, 80:146-159.

[17]Li LW, Ren XM, Guo FM, 2018. Modified multi-innovation stochastic gradient algorithm for Wiener–Hammerstein systems with backlash. J Franklin Inst, 355(9):4050-4075.

[18]Li LW, Zhang HL, Ren XM, 2020. A modified multi-innovation algorithm to turntable servo system identification. Circ Syst Signal Process, 39(9):4339-4353.

[19]Martin E, Lennart L, 2005. Linear approximations of nonlinear FIR systems for separable input processes. Automatica, 41(3):459-473.

[20]Mu BQ, Chen HF, 2014. Recursive identification of errors-in-variables Wiener–Hammerstein systems. Eur J Contr, 20(1):14-23.

[21]Mzyk G, Wachel P, 2017. Kernel-based identification of Wiener–Hammerstein system. Automatica, 83:275-281.

[22]Naitali A, Giri F, 2016. Wiener–Hammerstein system identification—an evolutionary approach. Int J Syst Sci, 47(1):45-61.

[23]Paduart J, Lauwers L, Pintelon R, et al., 2012. Identification of a Wiener–Hammerstein system using the polynomial nonlinear state space approach. Contr Eng Pract, 20(11):1133-1139.

[24]Piroddi L, Farina M, Lovera M, 2012. Black box model identification of nonlinear input-output models: a Wiener–Hammerstein benchmark. Contr Eng Pract, 20(11):1109-1118.

[25]Rijlaarsdam D, Oomen T, Nuij P, et al., 2012. Uniquely connecting frequency domain representations of given order polynomial Wiener–Hammerstein systems. Automatica, 48(9):2381-2384.

[26]Ross S, 2014. Introduction to Probability Models (11^th Ed.). Elsevier, Amsterdam, the Netherlands.

[27]Shaikh MAH, Barbé K, 2019. Wiener–Hammerstein system identification: a fast approach through Spearman correlation. IEEE Trans Instrum Meas, 68(5):1628-1636.

[28]Sjöberg J, Schoukens J, 2012. Initializing Wiener–Hammerstein models based on partitioning of the best linear approximation. Automatica, 48(2):353-359.

[29]Sjöberg J, Lauwers L, Schoukens J, 2012. Identification of Wiener–Hammerstein models: two algorithms based on the best split of a linear model applied to the SYSID’09 benchmark problem. Contr Eng Pract, 20(11):1119-1125.

[30]Škrjanc I, 2021. An evolving concept in the identification of an interval fuzzy model of Wiener–Hammerstein nonlinear dynamic systems. Inform Sci, 581:73-87.

[31]Tiels K, Schoukens M, Schoukens J, 2014. Generation of initial estimates for Wiener–Hammerstein models via basis function expansions. IFAC Proc Vol, 47(3):481-486.

[32]Wang ZY, Zhang Y, Jin QB, et al., 2022. Wiener models robust identification of multi-rate process with time-varying delay using expectation-maximization algorithm. J Process Contr, 118:126-138.

[33]Weber D, Gühmann C, 2021. Non-autoregressive vs autoregressive neural networks for system identification. IFAC-PapersOnLine, 54(20):692-698.

[34]Zong TC, Li JH, Lu GP, 2021. Auxiliary model-based multi-innovation PSO identification for Wiener–Hammerstein systems with scarce measurements. Eng Appl Artif Intell, 106:104470.