Dependability for Systems with a Partitioned State Space

This work is concerned with the dependability characteristics of reliability models which have a partitioned state space. In applications, these partitions will correspond to various degrees of system performance. Both discrete and continuous time systems are considered as well as Markov and semi-Markov systems. With these models, one may develop many dependability characteristics including the number of working periods during an interval, the number of repair periods until system breakdown, and the total time spent in the set of working states. The author shows how the theory may be applied using numerous examples with three computing packages: MATLAB, MAPLE, and the NAG FORTRAN 77 subroutine library.