Reliability Allocation for Series‐Parallel Systems subject to Potential Propagated Failures

Abstract

To analyze the dependent failures in the early stage of system development, this paper considers the potential propagated failures in the reliability allocation process. Factors which can be used to not only measure the component importance but also to reflect the influence brought by propagated failures are proposed. Specifically, cooperative game theory is introduced to explore how the propagated failures affect the failure severity level. Failure rates are obtained by using the Alpha Factor Model with the consideration of dependence among components. Reliability improvement rate is also developed to proportionally assign the target improvement of system reliability to the corresponding components. Furthermore, reliability allocation frameworks for series, parallel and series‐parallel systems are designed respectively to make the proposed model meet a wide range of applications. An illustrative example of a hydraulic cooling system is presented to show how the proposed approach is applied. The allocation results demonstrate that the proposed method can achieve a valid reliability improvement with the minimum error.

Optimal allocation of reliability improvement target based on multiple correlation failures and risk uncertainty

“…The correlation failure is usually summarized as cascading failure, common cause failure, and negative correlation failure [2]. The cascading failure refers to the behavior that a component failure facilitates subsequently the failure of another component in a system, which has an effect of magnifying 𝜆 𝑖 of the latter and 𝑠 𝑖 of the former compared with the independent failure.…”

“…Yadav and Zhuang [32] proposed to allocate the system reliability improvement target to the subsystems. Cao et.al [2,3] developed an allocation weight calculation theory by improving and completing the Risk Priority Number (RPN) model. Chen et.al [4] improved the efficiency of obtaining weights via multi-criteria decision-making methods.…”

A behavior‐driven reliability modeling method for complex smart systems

“…For example, the phased‐mission behavior has recently been modeled using an extended object‐oriented Petri net considering failed task reexecution 30 and universal generating functions considering performance sharing and transmission loss 31 in the physical domain. The failure propagation behavior has recently been examined for critical infrastructure systems, 32 supply chains, 33 and series‐parallel systems 34 . The common‐cause failure behavior has been investigated using Bayesian networks 35 and a survival signature‐based method for satellite systems 36 .…”

Progressive system safety and reliability analysis: A sustainable game theory approach

“…[24][25][26][27][28][29][30][31][32][33][34][35][36][37] Three examples include the dependent failures at the early life of a system, analyzed by cooperative game theory. 38 Gao et al 18 have assessed the impact of decision-making errors on strategies of software detection. 27 In another study, urban pipeline accidents are analyzed with the integration of game theory approaches and safety engineering systems.…”