数据缺失情况下水淹天然气管道泄漏风险分析

发布时间：2018-11-21 15:55

【摘要】：面向数据缺失情况下水淹天然气管道泄漏风险量化分析的需求,提出一种基于贝叶斯网络(BN)和模糊集理论(FST)的概率风险分析方法。首先采用故障树分析(FTA)法分析水淹天然气管道泄漏失效致因,并映射得到相应的BN模型;然后针对基本事件失效概率数据缺失的情况,用专家知识引出概率,替代缺失的统计失效概率;为处理概率引出过程中专家知识的模糊性和主观性导致的不确定性,结合FST与多专家层次分析引出模糊概率,将其作为实际先验概率输入BN模型,进行量化分析。以某复线水淹天然气管道为例,应用所提方法分析其泄漏风险,结果表明:用该方法能够在数据缺失情况下表征并量化泄漏风险,同时BN的正向预测和概率更新能力可用来评估动态风险、识别关键失效因素。
[Abstract]:A probabilistic risk analysis method based on Bayesian network (BN) and fuzzy set theory (FST) is proposed for quantitative analysis of leakage risk of flooded natural gas pipeline in the absence of data. Firstly, the fault tree analysis (FTA) method is used to analyze the cause of leakage failure of waterflooded natural gas pipeline, and the corresponding BN model is obtained by mapping. Then, in view of the absence of the basic event failure probability data, the probability is derived by expert knowledge to replace the missing statistical failure probability. In order to deal with the uncertainty caused by the fuzziness and subjectivity of expert knowledge in the process of probabilistic extraction, the fuzzy probability is derived from FST and multi-expert hierarchy analysis, and the fuzzy probability is input into the BN model as a practical priori probability for quantitative analysis. Taking a waterflooded natural gas pipeline on a double line as an example, the leakage risk of the pipeline is analyzed by using the proposed method. The results show that the method can be used to characterize and quantify the leakage risk in the absence of data. At the same time, the forward prediction and probability updating ability of BN can be used to evaluate dynamic risk and identify key failure factors.
【作者单位】：中国石油大学(北京)机械与储运工程学院;代尔夫特理工大学技术政策与管理学院;
【基金】：国家自然科学基金资助(51005247) 中国石油大学(北京)优秀青年科研基金资助(2462015YQ0406) 中国国家留学基金资助(留金欧[2016]6188) 荷兰互换奖学金项目(12273)
【分类号】：TE88

【参考文献】