Abstract
To expand and to improve the ability of fuzzy control inference system, we have analyzed the characteristic of fuzzy control rules in the unstable system with multi-degrees of freedom using computer simulation. In this study we have managed to control the inverted double pendulum. From our experiences, we have adopted a two-stage inference method, in which first, we infer the force to control the each arm of the double pendulum independently, and then we unify and infer the actural value of the force from the values that was infered in the first step. To make the control system more flexible and more robust, we proposed a simple and strong adaptive control method of range scaling method in which the system reduces or expands the scale of the range of fuzzy variables dynamically. This method resolves the conflict of requests such as to make the inverted pendulum standibng with the wide range of the initial values and to control the pendulum more precisely. We could the inverted double pendulum using the fuzzy control rule sets and adapting the range scaling method by a computer simulation. The fuzzy control system reveals an accurate control and also achieves a robust control by the simple range scaling method proposed in this paper.
