Use of evidential reasoning for eliciting Bayesian subjective probabilities in human reliability analysis

Use of evidential reasoning for eliciting bayesian subjective probabilities in human reliability analysis: A maritime case

“…The rule base of modelling probabilistic causal relation between parent-child nodes in the BNbased CREAM (e.g. Figure 2) is developed (Yang et al, 2013;Abujaafar et al, 2016). It reflects the interaction among the nine CPCs originally defined in CREAM (Hollnagel, 1998).…”

“…For instance, the prospective assessment model of the basic approach to estimate human error probability (HEP) in CREAM (Hollnagel, 1998) cannot provide a crisp value of the consequences of human performance, and the HEP estimation mechanism is not sensitive to minor changes associated with the nine common performance conditions (CPCs) in CREAM (Yang et al, 2013;Xi et al, 2017). A fuzzy Bayesian reasoning approach was developed to deal with this problem through using Bayesian Networks (BNs) to model the parent-child relationship between the CPCs and Contextual Control Model Controlling Modes (COCOM-CMs) in CREAM (Yang et al, 2013;Abujaafar et al, 2016). However, it requires too much information about the prior conditional probabilities assigned to the node of COCOM-CMs, jeopardising the applicability of the approach.…”

Prediction Model of Human Error Probability in Autonomous Cargo Ships

“…(1) Deburring In this paper, the mean area method is used to process the fuzzy probability and obtain the exact probability [30] . The formula is shown as follows: (7) (2) Probability normalization For each basic event, the sum of the state probabilities must be equal to one, therefore, the probability given by ( 4) should be normalized as follows: [31][32] (8)…”