Identification of the Fuzzy Hyperbolic Model
Birkhäuser Boston eBooks · 2007
The two types of fuzzy models used frequently in the area of fuzzy control 4 are the fuzzy relation model and the Takagi-Sugeno (T-S) fuzzy model. The main dis-advantages of these models are summarized as follows. (1) The premise structures have to be identified every time when a fuzzy model is updated on-line and heavy computation burden will be encountered. (2) The model cannot easily be used for controller design. This is because each rule in the above fuzzy models is a local model. Because the output from each model is only a local output, local models are not true dynamic models. By combining these outputs, an overall input-output model can be developed and used for stability analysis. However, in this case, it is not possible to design a feedback controller using state-space analysis. (3) The identification of these fuzzy models requires long convergence time and they are easily trapped in local minimum points during identification.