DESIGN OF A MODEL-BASED DIAGNOSIS SYSTEM FOR A THREE-TANK PLANT USING POSSIBLE CONFLICTS

Continuous systems may exhibit disastrous consequences due to different kinds of faults. To reduce its consequences, systems must detect and identify these faults. The fault detection and isolation tasks can be carried out by means of model-based techniques. Among the model-based diagnosis techniques, consistency-based approach is the most used within the Artificial Intelligence community. In this paper, we present the application of a consistency-based diagnosis technique, called Possible Conflicts, to a laboratory plant. Possible Conflicts can be computed using a specific software tool. The system and the Possible Conflicts have been implemented using the simulation software tool SIMULINK©. Finally, the system has been tested in a simulated benchmark. Index Terms: FDI, model-based diagnosis