Diagnosability Analysis using Quasi-Static Models

Diagnosability of faults relies on adequate sensors to provide redundancy relations so that the discrepancies of predictions and observations can be detected. This paper describes a model-based fault detection and diagnosis approach for dynamic systems. The approach uses the state space representation of a quasi static model to analyze fault discriminability. The methodology allows the identification of the set of sensors that guarantees discriminability of anticipated faults. The effectiveness of this approach has been evaluated through its application on a two-tank system