A New Method of Human Reliability Analysis Based on the Correlation Coefficient in the Evidence Theory and Analytic Hierarchy Process Method

Human reliability analysis (HRA) is a systematic technique for determining the impact that humans have on high-risk systems. Dependence assessment is a crucial aspect of HRA, as it quantifies the impact of an operator’s inability to accomplish one task on the failure probabilities of subsequent tasks. Many traditional methods for dependence assessment usually need analysts to make a definitive judgment all by themselves. It strongly relies on the knowledge and experience of domain specialists. To address this issue, a model based on the correlation coefficient in the Dempster–Shafer evidence theory and the analytic hierarchy process method is proposed in this paper. We use Dempster–Shafer evidence theory (D–S evidence theory) to model uncertainty and subjectivity in the assessment process. And the correlation coefficient is used to modify the basic probability assignment function to make the fusion formula get more accurate results. At the same time, the analytic hierarchy process is used to model the hierarchical relationship of multifactor decision-making problems. Benefiting from these functions and a united framework structure, the method has an extraordinary ability to reduce subjectivity and ambiguity in language evaluation. In the case study, the proposed method was used to analyze the probability of dependence between two human operations on actual nuclear power plant data. The proposed method is particularly suitable for dependency assessments that are intrinsically linguistic in nature.