A knowledge-based reliability analysis system
Martin O. Hofmann, Claire Lynne McCullough · 2002
Probabilistic reliability analysis is reviewed, and an expert system that assists a reliability engineer in performing risk analysis for complex systems is presented. The system releases the reliability engineer from the repetitive tasks of enumerating support states and modifying the primary failure equation and assists in organizing target system descriptions into an explicit model and in deriving a block diagram from it. The methodology used to derive Boolean failure equations from a description of the target system is described. System functionality is modeled in such a way that reliability analysis can be facilitated and automated as far as possible. Knowledge about objects in the application domain, i.e. components and functions, as well as knowledge about the analysis process is incorporated into the analysis system. Knowledge is represented in the form of frames and rules in the function dependency model and in the supporting agents.>