Simulation methods for system reliability using the survival signature

Abstract

Recently, the survival signature has been presented as a summary of the structure function which is sufficient for computation of common reliability metrics and has the crucial advantage that it can be applied to systems with components whose failure times are not exchangeable. The survival signature provides a huge reduction in required information, e.g. for its storage, compared to the full structure function, its implementation to larger systems is still difficult in a purely analytical manner and simulations may be required to derive the reliability metrics of interest. Hence, the main question addressed in this paper is whether or not the survival signature provides sufficient information for efficient simulation to derive the system’s failure time distribution. We answer this question in the affirmative by presenting two algorithms for survival signature-based simulation. In addition, we present a third simulation algorithm that can be used in case of repairable components. It turns out that these algorithms are very efficient, beyond the initial advantage of requiring only the survival signature to be available, instead of the full structure function.