Failure risk assessment by multi-state dynamic Bayesian network based on interval type-2 fuzzy sets and leaky-weighted sum algorithm: A case study of crude oil pipelines

• Published in: Expert systems with applications Vol. 250; p. 123942
• Main Authors: Liu, Jiawei, Li, Xiufeng, Zhang, Yixin, Li, Tao, Wei, Liping, Wu, Wei
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 15.09.2024
• Subjects: Dynamic Bayesian network; Interval type-2 fuzzy numbers; Interval type-2 fuzzy sets-based similarity aggregation method; Leaky-weighted sum algorithm; Risk assessment; Interval type-2 fuzzy sets-based similarity aggregation method; Leaky-weighted sum algorithm; Dynamic Bayesian network; Risk assessment; Interval type-2 fuzzy numbers
• ISSN: 0957-4174
• DOI: 10.1016/j.eswa.2024.123942

Failures of the pipelines can not only result in economic losses, but also potentially lead to serious safety accidents. Therefore, it is important to assess the failure risk of pipelines in order to prevent and mitigate pipeline failure accidents. This study proposed a method for failure risk assessment in process systems that combines the dynamic Bayesian network (DBN) with interval type-2 fuzzy sets (IT2FS). In this method, the IT2FS were applied to reduce the subjectivity and uncertainty of expert opinions and the bias between individual opinions. Specifically, an IT2FS-based similarity aggregation method (IT2FS-SAM) was introduced to collect and aggregate the prior probabilities of the parent nodes in the DBN, and an improved weighted sum algorithm (namely leaky-weighted sum algorithm, Leaky-WSA) was developed to obtain the conditional probability tables (CPTs) of the DBN, effectively reducing the number of expert opinions required. Finally, the feasibility of the method was demonstrated through the failure risk assessment of a crude oil gathering pipeline. Using prediction and backward inference with the DBN, the failure probability and main failure causes of the pipeline were determined, allowing pipeline managers to take appropriate preventive and maintenance measures.