Robust detection and isolation of unknown sensor faults using interval observers and invariant sets

Conventional interval observer-based sensor fault detection and isolation (FDI) approach mainly considers given magnitude fault. This paper proposes a novel robust detection and isolation approach of unknown sensor faults based on interval observers and invariant sets, which extends sensor faults with given magnitudes to sensor faults with unknown but bounded magnitudes. In addition, in order to reduce the conservativeness of proposed sensor-fault FDI approach, based on invariant sets, a collection of guaranteed FDI conditions are established from all interval observers. Under these conditions, all unknown but bounded faults can be detected and isolated. At the end, a numerical example is used to illustrate the effectiveness of the proposed approach.