A model-oriented fault diagnosis system for microprocessor-based systems

It is a difficult task to automatically diagnose the faults of microprocessor-based digital systems, since their functions have rapidly increased. In this paper, we’ve developed a hierarchical description scheme to model the system under diagnosis (SUD). During the stage of fault identification in our diagnosis system, pseudo-simulation plays a significant role in evaluating the expected behavior of the SUD. The features of using the pseudo-simulation approach are not only solving the cyclic problem, but also evaluating a smallest candidate set. During the stage of fault isolation, forward analysis is adopted to analyze the fault propagation path and to generate a plausible candidate set. In addition, an auxiliary test-assistant board has been designed to handle the test operations. Through the proposed fault diagnosis system, field maintenance and troubleshooting of microprocessor-based systems has become an easier task.