15th Triennial World Congress of the International Federation of Automatic Control
  Barcelona, 21–26 July 2002 
AN ADAPTIVE FUZZY CONTROLLER WITH CMAC-BASED SCALING FACTORS
Liu Chaoying*, Youichi Hirashima, Akira Inoue
Faculty of Engineering, Okayama Univ., 3-1-1 Tsushima-Naka, Okayama,
700-0082, Japan
* Hebei Univ. of Science and Technology, Shijiazhuang, 050018, China

In this paper, a new adaptive fuzzy controller with CMAC-based adaptive scaling factors is proposed. By using the proposed method, scaling factors can be tuned on-line, and control performance can be improved. In addition, CMAC is used only for tuning scaling factors, so that it is easy to implement and other fuzzy parameters can be designed by conventional method. Furthermore, robustness for modeling error can be improved as compared to conventional method. Finally, simulation results are shown to demonstrate that the proposed method has better dynamic and static property and has better robustness than conventional method.
Keywords: adaptive fuzzy control, fuzzy hybrid system, CMAC, fuzzy parameter tuning, on-line learning
Session slot T-Tu-A04: Fuzzy adaptive control/Area code 3e : Fuzzy and Neural Systems