Selective maintenance optimization for mission-oriented dynamic dependent systems under uncertain future operating environments

Various selective maintenance (SM) models have been proposed to address the maintenance optimization problem for mission-oriented dependent systems (MODS). However, the effectiveness of these models depends on two assumptions: (1) the system operates in a deterministic environment, and (2) the dependencies are unchanged over the life cycle. Considering that future operating environments are generally stochastic and that the stochastic-dependencies (s-dependencies) between components/subsystems could be dynamically affected by maintenance activities and operating environments, traditional models may not meet the demand for accurate reliability assessment in real engineering. Therefore, this paper proposes a novel SM and repairpersons assignment (SM-RA) model for mission-oriented dynamic dependent systems (MODDS). In the proposed model, probabilistically occurring operating environments alter system reliability by directly impacting the failure models of sensitive components and indirectly influencing those of non-sensitive components. Subsequently, multilevel s-dependencies between components/subsystems in the MODDS are modeled by employing hierarchical Archimedean copulas (HACs). The scenario-based dependency parameters for each level in the copula function after maintenance are dynamically updated using multi-stage maximum likelihood estimation (MLE). Finally, a differential evolution (DE) algorithm is designed to solve the optimization model, and the effectiveness of the proposed method is verified by numerical examples.