A reliability evaluation method for multi-state flow networks considering network configuration adjustment

Abstract

Reliability serves as a critical metric for assessing the performance of multi-state flow networks (MFNs). During the processes of designing, optimizing, or conducting resilience analysis on these networks, the configuration of the network might undergo changes. Consequently, the reliability evaluation process may need to be iterated multiple times to identify an acceptable network configuration in these scenarios. However, due to the inherent NP-hard complexity of reliability assessment in MFNs, repeated evaluations can lead to high computational costs. To address this challenge, we propose a comprehensive MFN reliability evaluation framework that enhances both efficiency and accuracy when network configurations adjust. This framework integrates three key methods, each designed for different scenarios. An improved state space decomposition (SSD) method, and a deeper SSD method are proposed to obtain the sets of state space of a MFN that can ensure the preset accuracy demand. Besides, we developed a state space reconstruction (SSR) method to update the sets of state space when network configuration changes. These methods work collaboratively within the framework to reduce computational costs while ensuring reliable assessments. Performance evaluations demonstrate that the proposed framework significantly improves computational efficiency and maintains high result accuracy, making it suitable for iterative MFN reliability assessments.