查看文章

Natural gas pipelines are susceptible to external and internal risk factors, such as corrosion, environmental conditions, external interferences, construction and design faults, and equipment failures. Bayesian Networks (BN) is a promising risk assessment approach widely used to evaluate these risk factors. One of BN's inherent limitations is its inability to accurately capture statistical dependencies and causal relationships, which can be overcome by incorporating expert elicitation into BN. To account for uncertainty and vagueness in assessing pipeline failure risks, fuzzy set theory (FST) can be combined with BN, commonly known as Fuzzy Bayesian Networks (FBN). This study developed an FBN framework that uses linguistic variables to calculate fuzzy probability (FPr) through domain expert elicitation, and crisp probabilities (CPr) are computed using historical incident data from the Pipeline and Hazardous …