Abstract: We propose a method of adaptive fuzzy observer design for a class of uncertain nonlinear systems. The observer uses fuzzy system to approximate nonlinear functions so that the structure and parameters of the observed object are not needed in the design. The estimation error, approximation error and external disturbance are supervised by a robust term to provide favorable observation performance. This observer is an extension for the nonlinear Luenberger adaptive observer. Effectiveness of the method is proven theoretically and verified in simulation.

Key words: fuzzy system, state observer, nonlinear system, adaptive