Vol 10, No 1 (2006) > Articles >

Modeling of Coupled-Tank System Using Fuzzy Takagi-Sugeno Model

Aries Subiantoro 1

Affiliations:

 

Abstract: This paper describes modeling of coupled tank system based on data measurement using fuzzy Takagi-Sugeno model. The fuzzy clustering method of Gustafson-Kessel algorithm is used to classify input-output data into several clusters based on distance similarity of a member of input-output data from center of cluster. The formed clusters are projected orthonormally into each linguistic variables of premise part to determine membership function of fuzzy Takagi-Sugeno model. By estimating data in each cluster, the consequent parameters of fuzzy Takagi-Sugeno model are calculated using weighted least-squares method. The resulted fuzzy Takagi-Sugeno model is validated by using model performance parameters variance-accounted-for (VAF) and root mean square (RMS) as performance indicators. The simulation results show that the fuzzy Takagi-Sugeno model is able to mimic nonlinear characteristic of coupled-tank system with good value of model performance indicators.
Keywords: System modeling, fuzzy Takagi-Sugeno, fuzzy clustering, coupled-tank
Published at: Vol 10, No 1 (2006) pages: 28-33
DOI:

Access Counter: 1004 views, 1356 PDF downloads, .

Full PDF Download

References:

R. Babuska, Fuzzy Modeling for Control, Kluwer Academic Publisher, Boston, 1998.

M. Sugeno, T. Yasukawa, IEEE Trans. on Fuzzy Systems 1 (1993) 7.

R. Babuska, H. Verbruggen, Proceeding European Control Conference, Rome, Italy, 1995, p. 1207.

K.J. Astrom, B. Wittenmark, Adaptive Control, Addison-Wesley, New York, 1989.

K. Narendra, J. Balakrishnan, M. Ciliz, IEEE Trans. on Control Systems, 15 (1995) 37.

L. Ljung, System Identification: Theory for the User, Prentice Hall, New Jersey, 1987.

D. Gustafson, W. Kessel, Proceeding IEEE CDC, San Diego, USA, 1979, p. 761.