Abstract
In a stochastic flow network (SFN), arcs and nodes are composed of multiple components, each of which can be in a binary state of success or failure, resulting in stochastic capacity states influenced by component failures. The reliability of components generally varies with time. This study aims to assess the time-varying system reliability of an SFN. Incorporating maintenance operations, such an SFN is termed a maintainable stochastic flow network (MSFN). Time-varying system reliability is defined as the probability that an MSFN can transmit a given demand from a source to a sink at a different time. This study includes an analysis of optimal maintenance timing for the components in the MSFN to maximize time-varying system reliability by developing a novel genetic algorithm. A benchmark example demonstrates the utility of the proposed genetic algorithm. Two practical cases further validate its efficiency and effectiveness through comparative analysis and management insights.