Unknown input observer design for fuzzy systems with uncertainties

This paper investigates the problem of unknown input observer design for both discrete and continuous-time T-S fuzzy systems with uncertainties. After doing appropriate processing to the model and reasonable analysis to the error expression of the system, the observer design conditions are proposed in LMI form based on Lyapunov theory. More important is the introduction of a new decoupling method which can further reduce the conservatism. The idea can eliminate the influence of the unknown inputs, and guarantee the error of the state estimation is bounded when the uncertainties are nonzero. Finally, an appropriate example is given to show the effectiveness of the algorithm, especially the excellent estimate ability of the observer in initial time.