Reliability Assessment of Reconfigurable <i>k</i> -out-of- <i>n</i> Systems With Functional Dependency

The <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">k</i>-out-of-<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">n</i> system with functional dependency (FDEP), as a typical structure, has widespread applications in a diversity of engineered system. These systems are characterized by components that perform distinct functions, and are connected through flexible intercomponential support relations. This flexibility allows for dynamic adjustment of the support strategy in response to component failures, achieved through connections between components’ interfaces or controlled by additional components, such as valves and switches. Even though previous article has demonstrated effectiveness in assessing reliability of <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">k</i>-out-of-<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">n</i> systems with FDEP, it often overlooks the essential investigation of flexible support relations among components, resulting in inaccurate system reliability assessment. To fill this research gap, this article introduces a novel framework that integrates a parameter time-varying discrete dynamic Bayesian network (PTVDDBN) and a tailored Hungarian algorithm with a depth-first search (DFS) strategy, namely the PTVDDBN–HDFS method, to advance reliability assessment of <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">k</i>-out-of-<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">n</i> systems with flexible support relations. Specifically, the PTVDDBN-based architecture captures the system's stochastic degradation over time, and its components’ lifetime could follow an arbitrary probability distribution. From a graph set-based perspective, the support strategy designated in the system is dynamically adjusted via the DFS strategy. The optimal system performance under various component state combinations is further converted to conditional probability table parameters within the PTVDDBN model. A practical case study of a kerosene filling system at a space launch site is showcased to illustrate the application and effectiveness of the PTVDDBN–HDFS method.