A model-based approach to fault diagnosis of FMS

The success of an FMS is very much dependent on its trouble-free operation. In an FMS, it is therefore important to identify the root causes and take corrective actions as soon as a failure or fault symptom has been detected in any of the system components. Traditionally, fault diagnosis of FMS is performed by humans and the process is time-consuming. However, the process can be automated to improve time-effectiveness. This paper proposes a model-based approach for automating the process. The approach utilizes a fuzzy digraph coupled with worst-first search reasoning for tracing root causes of a system failure. In this paper, the approach is explained and illustrated with an example