Fuzzy bayesian network based on an improved similarity aggregation method for risk assessment of storage tank accident

Fuzzy Bayesian network (FBN) has been widely used for risk assessment of accidents in process industries to deal with complex causality and uncertainty arise from complex interdependence among risk factors, insufficient data and complex environment. The similarity aggregation method (SAM) is a method of aggregating fuzzy opinions considering consensus degree. However, SAM does not take into account the deviations caused by individual differences in dealing with consistency, which will bring a certain degree of uncertainty. Therefore, this work proposes an improved SAM based FBN model to better deal with vari-ous types of uncertainty. This methodology makes the prediction results of the storage tank accident more accurate and reliable. The result analysis indicates that the improved SAM is of significance to improving the reliability of the input data of FBN. Then, the critical analysis of the root node shows the effectiveness and reliability of FBN in identifying the critical events of the storage tank accident. The proposed method can predict the probability of storage tank accidents, determine the proportion of main contributing fac-tors and identify the critical causes of storage tank accidents more reliably and accurately. It can provide important supporting information for decision makers to optimize risk management strategies. (C) 2020 Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.