A Novel Dependence Assessment Method With Comparative Linguistic Expression and Hybrid Cloud Model in Human Reliability Analysis

Human reliability analysis (HRA) offers a framework for the identification and evaluation of human errors in large-scale industries like civil aviation. Dependence assessment plays a critical role in HRA, which is to evaluate the dependence degrees among human error events (HFEs). Dependence assessment necessitates expertise and knowledge from experts, however, due to the complexity of real-world decision-making, there inevitably encounters various uncertainties when experts assess the dependence between HFEs. Considering the above issues, this paper introduces a novel method using comparative linguistic expression and hybrid cloud model with the help of the technique for human error rate prediction (THERP) to address dependence assessment in HRA. This paper conducts the comparative linguistic expression to capture the multi-presentation linguistic opinions from experts and develops a cloud transformation framework that utilizes the hybrid cloud models to represent and handle experts' opinions. Furthermore, two objective weight calculation approaches are proposed to determine the weights of the influential factors and experts without prior known weight information. The dependence degrees between HFEs can be obtained according to the cloud model and the THERP method. Finally, an empirical dependence assessment for HFEs in air traffic control (ATC) demonstrates the rationality and effectiveness of our proposed method. It can be concluded that our proposed method offers an applicable and effective way for dependence assessment in HRA.