Observer-based variable universe fuzzy control with dynamic self-structure

An improved observer-based robust adaptive variable universe fuzzy control is proposed in this paper. The structure of fuzzy system is on-line tuned by the dynamic self-structure method except scaling gains adjusted by adaptive laws. The global stability of the resulting closed-loop system is also guaranteed. The proposed controller has not only simplified design for the initial structure of fuzzy system compared with the original controller, but also less fuzzy approximation error with dynamic structure. First, the principle of the observer-based robust adaptive variable universe fuzzy control is presented briefly. Then, the method of dynamic self-structure is given. The proposed improved variable universe fuzzy control is applied to the control of the Duffing chaotic system. The feasibility for practical application is confirmed by simulation results.