Diagnosis of Uncertain Systems with Arbitrary Relative Degree and Unknown Input

Abstract

The problem of detecting sensor and actuator faults for a class of multivariable systems with arbitrary relative degree is considered. The proposed solution is based on the synthesis of unknown input observers. A modification is proposed that ensures a predefined convergence time, which is based on the dynamic regressor extension and mixing method and the Kreisselmeier scheme. The obtained solution is extended to the class of nonlinear systems with parametric uncertainties. Simulation results are provided to illustrate the effectiveness of the proposed approach.