Component uncertainty importance measure in complex multi-state system considering epistemic uncertainties
Rentong Chen, Shaoping Wang, Chao Zhang, Hongyan Dui, Yuwei Zhang, Yadong Zhang, Li Yang
Abstract
Importance measures can be used to identify the vulnerable components in an aviation system at the early design stage. However, due to lack of knowledge or less available information on the component or system, the epistemic uncertainties may be one of the challenging issues in importance evaluation. In addition, the properties of the aircraft system, which are the fundamentals of the component importance measure, including the hierarchy, dependency, randomness, and uncertainty, should be taken into consideration. To solve these problems, this paper proposes the component Uncertainty Integrated Importance Measure (component UIIM) which considers multiple epistemic uncertainties in the complex multi-state systems. The degradation process for the components is described by a Markov model, and the system reliability model is developed using the Markov hierarchal evidential network. The concept of integrated importance measure is then extended into component UIIM to evaluate the component criticality rather than the component state change criticality, from the perspective of system performance. A case study on displacement compensation hydraulic system is presented to show the effectiveness of the proposed uncertainty importance measure. The results show that the component UIIM can be an effective method for evaluating the component criticality from system performance perspective at the system early design.