Performance Margin and Reliability Modeling Method for Multi-Level Redundant System

This study proposes a multi-level performance margin modeling and belief reliability framework for redundant systems. Starting from system performance, a “performance–margin–reliability” linkage is established by defining the performance and margin of multi-level redundant systems and deriving performance, margin, and metric equations that account for failures. For complex redundant systems, a hierarchical Behavior Interaction Priority (BIP) modeling approach is developed to explicitly represent the normal and failure states of atomic component models. The effects of redundant components on the overall system are transformed into variations of performance parameters, enabling quantitative analysis of redundancy mechanisms. This paper proposes a boundary search algorithm for pruning optimization, which breaks through the computational bottleneck of non-analytic threshold sets in high-dimensional topological spaces. A case study on a power supply system with multi-level structural redundancy is conducted. Based on the proposed method, a performance-margin model of the redundant power supply system is constructed, critical states are analyzed, and system reliability is calculated. The results verify the effectiveness of the proposed margin-equation formulation and solution algorithm, offering practical guidance for reliability design of redundant systems.