Testability Evaluation of the Systems with Multi-outcome Imperfect Tests

摘要

In the fault diagnosis systems, the measured value of a test is generally divided into several intervals to isolate the faults. Furthermore, in the realistic systems, the value may be distributed over different intervals with a probability if a fault occurs, which means that the tests are multi-outcome imperfect tests. In this paper, a calculation method of metrics of fault detection and fault isolation based on extended dependency matrix is proposed, which can be applied to evaluate the testability of a system and consequently play a role in the design process of diagnosis systems.