Hierarchical models for systems reliability, maintainability, and availability

In this paper, the development of low-cost methodologies for the prediction and evaluation of system reliability, maintainability, and availability (RMA) is considered. Generalized Stochastic Petri Nets (GSPN's) are used as an analytical modeling tool. A general method of analysis for GSPN's by state aggregation is presented. This concept of aggregation is then applied to subnetworks, at the GSPN level, to reduce the model complexity. Using the above, a hierarchical model is obtained, based on the notion of time scale decompsotion, for system RMA. The above concepts are illustrated by examples.