Reliability optimization of parallel-series and series-parallel systems with statistically dependent components

Abstract

In this manuscript, we adopt a copula-based hierarchical dependent approach to analyze the redundancy allocation policy and statistical dependence effect on the reliability of two-parallel-series and series-parallel systems with interdependent subsystems from statistically dependent components and use a Schur-concave copula to model the statistical dependence between the outer subsystems, and an Archimedean copula to characterize the statistical dependence among the inner components. Firstly, we analyze and compare the reliability of different redundancy strategies for a two-parallel-series system in the sense of the usual stochastic order under the number of chosen components that are fixed and randomized. Secondly, we also provide sufficient conditions on the usual stochastic order of two different assembly policies for a two-series-parallel system when the number of chosen components is deterministic and randomized. These results can be conductive to allocate the two-parallel-series and two-series-parallel systems with interdependent subsystems from statistically dependent components in practical scenarios. Finally, some applications of the real-data and the potential managerial implications and future directions are also provided to enhance the system’s reliability.