抄録
This paper deals with a self-tuning fuzzy controller. The fuzzy controller is constructed with linguistic control rules which consist of the fuzzy variables and the fuzzy sets. Each of fuzzy sets is characterized by a membership function. The self-tuning fuzzy controller has a function that the parameters of membership functions are adequately adjusted by two learning rule sets.
One is the repeated learning rule set which adjusts the parameters of membership functions of the fuzzy controller after evaluating the control results with respect to the control performances. The other is the real time learning rule set which adjusts the parameters of the fuzzy controller by evaluating the control response at the real time.
By control simulation, it is shown that the desired control results are obtained and that the usefulness of the self-tuning fuzzy controller is recognized.
