Adaptive Observer Design with Unknown Inputs for Nonlinear Systems

Abstract:

For the nonlinear systems with unknown inputs and measurable noises, the design method of unknown input observer based on the TS model is discussed. At first, the calculation method of the coefficient matrix of the TS model constructed based on nonlinear system is presented. Then, the TS model is used as the design model, and the system is extended to an augmented system by extending the measurable noise as an auxiliary state. And then, for the TS model, a fuzzy adaptive observer is designed to achieve the objective of the simultaneous estimations of the states, the unknown inputs and the measurable noises of the nonlinear system. Meanwhile, the sufficient condition of the existence of the observer is given based on a linear matrix inequality. Finally, simulation examples demonstrate the effectiveness of this method.

Key words:adaptive observer; T-S model; unknown input reconstruction; measurement noise reconstruction; unmeasured premise variables