Using design information to support model-based fault diagnosis tasks

Abstract

In today’s complex systems, it can be difficult to specify the cause of a fault. In general, backward reasoning that derives the cause from the fault is used for diagnosis. Knowledge that shows the state of the target object based on the fault is necessary for this inference. However, especially in the field of space, each system is peculiar to each project, and knowledge that associates causes with faults tends to be insufficient. Model-based diagnosis is therefore used. It uses a model of a target system to recreate the fault, thus showing the condition of the target system. To construct such models, knowledge of the target system is required. We have developed a system that collects knowledge during the design process, and provides it to the model construction process.