Unknown input observer synthesis for discrete-time T-S fuzzy singular systems with application to actuator fault estimation

This work studies an unknown input observer (UIO) design dedicated to the Takagi-Sugeno fuzzy-approximation-based nonlinear singular systems with unknown inputs in the discrete-time domain. Not only the input equation, but also the output equation are assumed to be affected by unknown inputs. By restoring to a fuzzy Lyapunov function together with the well-developed linear matrix inequality (LMI) technique, some new relaxed sufficient criteria for the convergence of the designed UIO are exactly given and accurately proved. Significantly, consider that the actuator fault may occur frequently, which may result in a disastrous consequence for modern control systems. Therefore, an extended UIO called unknown input proportional integral observer is further investigated to reconstruct the actuator fault. All the design procedures are formulated in terms of LMIs. Finally, simulation studies on a numerical example as well as a practical one are provided to test the performance of the developed two procedures.