Dependence assessment in human reliability analysis based on evidence credibility decay model and IOWA operator

Dependence assessment among human errors plays an important role in human reliability analysis. The technique for human error rate is one of the most commonly used method for the assessment of dependence among HRA. However, most of the proposed method considered that the time factor related to the major factor “Closeness in Time” in the THERP can be elicited from experts' judgments, which is not appropriate for the time interval between two tasks is an accurate value can be measured. In this paper, we proposed a method based on the Evidence Credibility Decay Model (ECDM) and the Induced OWA operator mainly to assess the dependence between tasks in HRA which can also deal with the subjective limitations on the “CT”. Based on the ECDM-IOWA model, we can easily get the discount rate relates to the anchor points of “CT”, then consider the factor “Similarity of Performers” is changing by the time, the BBA of “SP” is discounted by the discount rate. Finally, the BBA of the reconstructed “SP” and other factors are fused to calculate the CHEP. Experimental results demonstrate that the proposed model not only represents the experts evaluate in a more objective way but also shows how the input factors and the parameter α in IOWA influence the dependence level.