Fault Diagnosis in the Presence of Parametric Variations

Abstract In this paper, we tackle the problem of fault detection and isolation in the presence of parametric variations. Under the influence of such parametric uncertainties, we want to detect and isolate external failures such as sensor and actuator faults. In particular, we deal with the case where the unknown parameters decoupling approach is not feasible for a reliable diagnosis scheme. Different ways to solve the problem are followed and discussed. We emphasize on a linearizing approach based on a two steps identification procedure of faults and parametric variations, and also on a direct nonlinear estimation approach.