| |
Automated diagnosis deals with techniques to determine the cause of any abnormal or unexpected behavior of physical systems. The key issue is that inferences have to be made from the limited sensor information available from the system. Some major categories of diagnostic technologies are rule-based systems, case-based reasoning systems, data-drive learning systems, and model-based reasoning systems among others. In this paper we will briefly introduce these categories and then focus on model-based reasoning. We will present the Hybrid Diagnosis Engine (HyDE) developed at the NASA Ames Research Center and its application to real problems.
|
|
