MC-reduction simulation approach with heuristic rules for reliability estimation in a multi-state flow network

Abstract

The system reliability of a multistate flow network (MSFN) holds significant importance as a critical indicator for assessing the probability of successfully transmitting a given demand. It is recognized that evaluating the system reliability of an MSFN presents a challenge due to its NP-hard property. Hence, the paper introduces a novel simulation approach based on minimal cuts (MCs) to estimate the system reliability of an MSFN. The MC-based simulation approach not only enhances computational efficiency during the simulation but also introduces two heuristic rules to remove the MCs that are likely to satisfy the demand. The first rule revolves around the concept of cut size, aiming to reduce the number of the MCs. Subsequently, a saturation threshold is implemented to preserve saturated MCs. By incorporating these two rules, the need to consider all MCs is circumvented, leading to notable gains in computational efficiency during the simulation. Furthermore, the proposed simulation approach effectively deals with non-integer demands directly for the system reliability estimation. Illustrative examples demonstrate the effectiveness and efficiency of the proposed MC-reduction simulation approach. Overall, the reduction in the number of MCs directly translates into a reduction in the time required, underlining the practical benefits of this innovative approach.