AUTHORS: Jessica A. Santos, Ginalber L. O. Serra

Download as PDF

ABSTRACT: This paper presents a novel methodology for evolving fuzzy identification of nonlinear systems in state space based on Hammerstein models. The nonlinear static characteristic is approximated by an evolving TakagiSugeno fuzzy model and the linear dynamics by a state space model. The recursive estimation of the linear model in state space is performed based on the system Markov parameters applied to the algorithm of minimum realization ERA. Computational results illustrate the effectiveness of the proposed method in the online identification of nonlinear systems

KEYWORDS: Evolving fuzzy systems, Hammerstein models, state space systems, Markov parameters

REFERENCES:

[1] Q. Jin, H. Wang, Q. Su, B. Jiang and Q. Liu, A novel optimization algorithm for MIMO Hammerstein model identification under heavy-tailed noise. ISA transactions, 2017.

[2] L. Shang, J. Liu and Y. Zhang, Recursive Fault Detection and Identification for Time-Varying Processes. Industrial & Engineering Chemistry Research 55.46 (2016): 12149-12160.

[3] E. Costa, and G. Serra, Swarm optimization based adaptive fuzzy control design from robust stability criteria. Journal of Intelligent & Fuzzy Systems 32.3 (2017): 1787-1804.

[4] S. Rodrigues Junior and G. Serra, A novel intel- ´ ligent approach for state space evolving forecasting of seasonal time series. Engineering Applications of Artificial Intelligence 64 (2017): 272- 285.

[5] M. Jafari, M. Salimifard and M. Dehghani, Identification of multivariable nonlinear systems in the presence of colored noises using iterative hierarchical least squares algorithm. ISA transactions 53.4 (2014): 1243-1252.

[6] W. Zhao, H. Chen and R. Tempo, Recursive Nonparametric Identification of Nonlinear Systems with Adaptive Binary Sensors. IEEE Transactions on Automatic Control (2017).

[7] R. Rattanawaorahirunkul, P. Sanposh and C. Panjapornpon, Nonlinear system identification of pH process using Hammerstein-Wiener model. Electronics, Information, and Communications (ICEIC), 2016 International Conference on. IEEE, 2016.

[8] S. Lajnef, N. Boulejfen and F. Ghannouchi, Cartesian Augmented Hammerstein Model for Nonlinearity and I/Q Impairments Compensation in Concurrent Dual-Band Transmitters. IET Communications (2017).

[9] Jalaleddini, Kian, and Robert E. Kearney, Subspace identification of SISO Hammerstein systems: application to stretch reflex identification. IEEE Transactions on Biomedical Engineering 60.10 (2013): 2725-2734.

[10] S. Tsai, and Y. Chen, A Novel Fuzzy Identification Method Based on Ant Colony Optimization Algorithm. IEEE Access 4 (2016): 3747-3756.

[11] E. Lughofer, C. Cernuda and M. Pratama, Generalized flexible fuzzy inference systems. Machine Learning and Applications (ICMLA), 2013 12th International Conference on. Vol. 2. IEEE, 2013.

[12] K. Hammar, T. Djamah and M. Bettayeb, Fractional Hammerstein system identification using polynomial non-linear state space model. Control, Engineering & Information Technology (CEIT), 2015 3rd International Conference on. IEEE, 2015.

[13] L. Jia, X. Li and M-S. Chiu. The identification of neuro-fuzzy based MIMO Hammerstein model with separable input signals. Neurocomputing 174 (2016): 530-541.

[14] E. Lughofer, FLEXFIS: A robust incremental learning approach for evolving Takagi–Sugeno fuzzy models. IEEE Transactions on fuzzy systems 16.6 (2008): 1393-1410.

[15] E. Lughofer, Evolving fuzzy systemsmethodologies, advanced concepts and applications. Vol. 53. Berlin: Springer, 2011.

[16] C-Y. Wu, J.S-H. Tsai, S-M. Guo, L-S. Shieh, J. Canelon, F. Ebrahimzadeh and L. Wang, A novel on-line observer/Kalman filter identification method and its application to inputconstrained active fault-tolerant tracker design for unknown stochastic systems. Journal of the Franklin Institute 352.3 (2015): 1119-1151.

[17] J-N. Juang and M. Phan. Identification and control of mechanical systems. Cambridge University Press, 2001.

[18] L. Ljung, System identification. Signal analysis and prediction. Birkhauser, Boston, MA, 1998. ¨ 163-173.

[19] J-N. Juang, M. Phan, L. Horta, R. Longman, Identification of observer/Kalman filter Markov parameters: theory and experiments. Journal of Guidance, Control, and Dynamics 16.2 (1993): 320-329.