A TS Fuzzy Model Derived from a Typical Multi-Layer Perceptron

Document Type: Research Paper


1 System Engineering and Mechatronics Group, Faculty of New Sciences and Technologies, University of Tehran, Tehran, Iran

2 Control and Intelligent Processing Center of Excellence, School of Electrical and Computer Engineering, University of Tehran, Tehran, Iran

3 System Engineering and Mechatronics Group, Faculty of New Sci- ences and Technologies, University of Tehran, Tehran, Iran


In this paper, we introduce a Takagi-Sugeno (TS) fuzzy model which is derived from a typical Multi-Layer Perceptron Neural Network (MLP NN). At first, it is shown that the considered MLP NN can be interpreted as a variety of TS fuzzy model. It is discussed that the utilized Membership Function (MF) in such TS fuzzy model, despite its flexible structure, has some major restrictions. After modifying the MF, we introduce a TS fuzzy model whose MFs are tunable near and far from focal points, separately. To identify such TS fuzzy model, an incremental learning algorithm, based on an efficient space partitioning technique, is proposed. Through an illustrative example, the methodology of the learning algorithm is explained. Next, through two case studies: approximation of a nonlinear function for a sun sensor and identification of a pH neutralization process, the superiority of the introduced TS fuzzy model in comparison to some other TS fuzzy models and MLP NN is shown.


[1] P. Angelov and X. Zhou, On line learning fuzzy rule-based system structure from data
streams, In 2008 IEEE International Conference on Fuzzy Systems within the IEEE World
Congress on Computational Intelligence, Hong Kong., (2008), 915-922.

[2] J. F. Baldwin and S. B. Kandarake, Asymmetric triangular fuzzy sets for classi cation model,
In Lecture Notesin Arti cial Intelligent, (2003), 364-370.
[3] J. N. Choi, S. K. Oh and W. Pedrycz., Identi cation of fuzzy models using a successive tuning
method with a variant identi cation ratio, Fuzzy Sets and Systems, 159 (2008), 2873-2889.
[4] B. L. R. De Moor, ed., DaISy: Database for the Identi cation of Systems, De-
partment of Electrical Engineering, 2008, ESAT/SISTA, K.U.Leuven, Belgium, URL:
http://homes.esat.kuleuven.be/~smc/daisy/. (visited on Oct. 10, 2010).
[5] H. Du and N. Zhang, Application of evolving Takagi-Sugeno fuzzy model to nonlinear system
identi cation, Applied soft computing, 8 (2007), 676-686.
[6] A. Fiordaliso, A constrained Takagi-Sugeno fuzzy system that allows for better interpretation
and analysis, Fuzzy Sets and Systems, 118 (2001), 307-318.
[7] M. Hell, S. P. Campinas, R. Ballini, Jr. P. Costa and F. Gomid, Training neurofuzzy networks
with participatory learning, In: proceeding of 5th Conference of the EUSFLAT, (2007), 231-
236, .
[8] J. S. R. Jang, ANFIS: Adaptive-network-based fuzzy inference system, IEEE Trans. Syst.
Man and Cybern, 23 (1993), 665-685.
[9] A. Kalhor, B. N. Araabi and C. Lucas, A new systematic design for habitually linear evolving
TS fuzzy model, Journal of Expert systems with applications, 39 (2012), 1725-1736.
[10] A. Kalhor, B. N. Araabi and C. Lucas, An online predictor model as adaptive habitually
linear and transiently nonlinear model, Evolving Systems, 1 (2010), 29-41.
[11] A. Kalhor, B. N. Araabi and C. Lucas, A new high-order Takagi-Sugeno fuzzy model based on
deformed linear models, Amirkabir Int. J. of Modeling Identi cation, Simulation and Control,
42 (2010), 43-52.
[12] A. Kalhor, B. N. Araabi and C. Lucas, Reducing the number of local linear models in neuro{
fuzzy modeling: A split-and-merge clustering approach, Applied Soft Computing Journal, 11
(2011), 5582{5589.
[13] N. Kasabov, DENFIS: Dynamic Evolving Neural-Fuzzy Inference System and its application
for time-series prediction, IEEE Trans. Fuzzy Syst., 10 (2012) 144-154.
[14] V. Krkov: Kolmogorov's theorem and multilayer neural networks, Neural networks 5: 501-
506, 1992.
[15] D. H. Lee, Y. H. Joo and M. H. Tak, Local stability analysis of continuous-time Takagi{
Sugeno fuzzy systems: A fuzzy Lyapunov function approach, Information Sciences, 257
(2014), 163-175.
[16] C. H. Lee and H. Y. Pan, Performance enhancement for neural fuzzy systems using asym-
metric membership functions, Fuzzy Sets and Systems, 160 (2009), 949-971.
[17] C. H. Lee and C. C. Teng, Fine tuning of membership functions for fuzzy neural systems,
Asian J. Control, 3 (2001), 216-225.
[18] G. Leng, T. M. Mc Ginnity and G. Prasad, An approach for on-line extraction of fuzzy rules
using a self-organising fuzzy neural network, Fuzzy Sets and Systems, 150 (2005), 211-243.
[19] C. Li C, K. H. Cheng and J. D. Lee, Hybrid learning neuro fuzzy approach for complex
modeling using asymmetric fuzzy sets, In: Proc. of the17th IEEE International Conf. on
Tools with Arti cial Intelligence, (2005), 397-401.
[20] C. Li, J. Zhou, X. Xiang, Q. Li and X. An, T-S fuzzy model identi cation based on a novel
fuzzy c-regression model clustering algorithm, Engineering Applications of Arti cial Intelligence,
22 (2009), 646-653.
[21] C. J. Lin and W. H. Ho, An asymmetric-similarity-measure-based neural fuzzy inference
system, Fuzzy Sets and Systems, 152 (2005), 535{551.
[22] T. J. Mc Avoy, E. Hsu and S. Lowenthal, Dynamics of pH in controlled stirred tank reactor,
Ind. Eng. Chem. Process Des. Develop., 11 (1972), 71-78.
[23] O. Nelles, Nonlinear System Identi cation, In: New York: Springer, (2001), Section 13.3.1,
[24] P. Nikdel, M. Hosseinpour, M. A. Badamchizadeh and M. A. Akbari, Improved Takagi{
Sugeno fuzzy model-based control of
exible joint robot via Hybrid-Taguchi genetic algorithm,
Engineering Applications of Arti cial Intelligence, 33 (2014), 12-20.

[25] J. Park and I. W. Sandberg, Universal Approximation Using Radial-Basis-Function Net-
works, Neural Computation, 3 (1991), 246-257.
[26] K. B. Petersen and M. S. Pedersen, The matrix cookbook, http://matrixcookbook.com, Version:
Nov. 14, 2008.
[27] Y. Shi, R. Eberhart and Y. Chen, Implementation of evolutionary fuzzy systems, IEEE Trans.
on Fuzzy Sys., 7 (1999), 109-118.
[28] T. Takagi and M. Sugeno, Fuzzy identi cation of systems and its applications to modeling
and control, IEEE Trans. Syst., Man, and Cybern., 15 (1985), 116-132.
[29] D. Wang, C. Quek and G. S. Ng, MS-TSKfnn: Novel Takagi-Sugeno-Kang fuzzy neural
network using ART like clustering., In: proceeding of IEEE international joint conference on
Neural Networks., (2004), 2361-2366.
[30] X. Xie, L. Lin and S. Zhong, Process Takagi{Sugeno model: A novel approach for han-
dling continuous input and output functions and its application to time series prediction,
Knowledge-Based Systems, 63 (2014), 46-58.
[31] H. Ying, General SISO Takagi{Sugeno fuzzy systems with linear rule consequent are univer-
sal approximators, IEEE Trans. on Fuzzy Systems, 6 (1998), 582-587.