基于贝叶斯网络城市埋地燃气管线风险评价研究

发布时间：2018-06-11 14:50

本文选题：城市埋地燃气管线 + 风险评价　；参考：《首都经济贸易大学》2015年硕士论文

【摘要】：城市燃气管线是城市重要的公共基础设施,城市的生命线之一,因其所输送的物质具有易燃、易爆、有毒的特殊危险性,其安全问题不容忽视。有效的城市埋地燃气管线风险评价能够减少和预防城市埋地燃气管线泄漏事故的发生,保障城市运行安全。论文主要是基于贝叶斯网络模型对城市埋地燃气管线的泄漏失效可能性的评价。在对国内外燃气管线事故统计分析和城市埋地燃气管线泄漏事故树分析的基础上总结城市埋地燃气管线泄漏事故的主要影响因素,并以管线泄漏失效可能性为评价目标建立贝叶斯网络模型,以设计与施工因素、腐蚀因素、第三方破坏、自然灾害为模型的四个主要方面,具体到三级父节点;使用noisy—or门确定节点条件概率,利用贝叶斯网络结构具有的强大的处理不确定性问题的能力来合理地综合专家的不同意见以更新节点边缘概率,最后利用构建的贝叶斯网络模型定量计算北京市曹慧璐Ch1段埋地燃气管线泄漏失效可能性。采用贝叶斯网络模型进行城市埋地燃气管线泄漏失效可能性评价的关键在于贝叶斯网络模型的构建和专家意见的有效集结。论文基于国内外城市燃气管线泄漏事故统计分析和事故树分析,总结主要影响因素为第三方破坏、老化腐蚀、设备及操作不当、管线本身缺陷、地面沉降、自然因素、人为蓄意破坏、管线间相互影响破坏等,以此为根据构建贝叶斯网络模型的结构;利用noisy—or门和贝叶斯网络结构具有的强大的处理不确定性问题的能力确定节点条件概率和节点边缘概率,完成模型的构建;以该模型对北京市曹慧路Ch1段燃气输配管段腐蚀因素影响下泄漏失效可能性进行定量评价,评价结果显示该管段埋地燃气管线泄漏失效可能性概率值约为0.735,参照第7章失效可能性等级划分表,该管线对应的泄漏危险等级为IV级,对比评价结果表明:利用贝叶斯网络模型对城市埋地燃气管线进行泄漏失效可能性风险评价,能够对城市区域内的燃气管网系统的安全管理提供有效依据。
[Abstract]:Urban gas pipeline is one of the most important public infrastructure and one of the lifelines of the city. Because of the special danger of flammable explosive and toxic the safety problems can not be ignored. Effective risk assessment of urban buried gas pipeline can reduce and prevent the leakage accident of urban buried gas pipeline and ensure the safety of urban operation. This paper is mainly based on Bayesian network model to evaluate the possibility of leakage failure of urban buried gas pipeline. Based on the statistical analysis of the gas pipeline accidents at home and abroad and the analysis of the gas pipeline leakage tree in the city, the main influencing factors of the gas pipeline leakage accident in the city are summarized. Taking the possibility of pipeline leakage failure as the evaluation goal, the Bayesian network model is established, with design and construction factors, corrosion factors, third party damage and natural disasters as the four main aspects of the model. Noisy-or gate is used to determine the conditional probability of nodes, and the strong ability of Bayesian network structure to deal with uncertain problems is used to reasonably synthesize the different opinions of experts to update the edge probability of nodes. Finally, the probability of leakage failure of buried gas pipeline in Ch1 section of Cao Huilu in Beijing is calculated quantitatively by using the constructed Bayesian network model. The key to evaluate the leakage failure possibility of urban buried gas pipeline by using Bayesian network model lies in the construction of Bayesian network model and the effective assembly of expert opinions. Based on the statistical analysis and accident tree analysis of leakage accidents of urban gas pipelines at home and abroad, the paper summarizes that the main influencing factors are third-party destruction, aging and corrosion, improper equipment and operation, pipeline defects, land subsidence, natural factors, etc. The structure of Bayesian network model is constructed based on artificial deliberate destruction and pipeline interaction failure. Using the powerful ability of noisy-or gate and Bayesian network structure to deal with uncertain problems, the conditional probability and edge probability of nodes are determined, and the model is constructed. Based on the model, the possibility of leakage failure was quantitatively evaluated under the influence of corrosion factors in Ch1 section of Caohui Road, Beijing. The evaluation results show that the probability of leakage failure of the buried gas pipeline is about 0.735. According to Chapter 7, the leakage risk level of the pipeline is grade IV. The comparison and evaluation results show that the risk assessment of leakage failure probability of urban buried gas pipeline based on Bayesian network model can provide an effective basis for the safety management of gas pipeline network system in urban area.
【学位授予单位】：首都经济贸易大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TU996.9

【参考文献】