岳帅帅, 范文换, 王文,等.基于可拓理论的气煤叠置区天然气管道安全风险评价研究[J].河南理工大学学报（自然科学版）,2026,45(2):195-203.

YUE S S, FAN W H, WANG W,et al.Study on safety risk assessment of natural gas pipelines in gas-coal overlap zones based on extension theory[J].Journal of Henan Polytechnic University(Natural Science) ,2026,45(2):195-203.

基于可拓理论的气煤叠置区天然气管道安全风险评价研究

岳帅帅1, 范文换1, 王文1,2, 郭润生3, 芦晓伟4

1.河南理工大学 能源科学与工程学院，河南 焦作  454003；2.煤炭安全生产与清洁高效利用省部共建协同创新中心，河南 焦作  454003；3.中国石油化工有限公司 华北油气分公司，河南 郑州  450000；4.中国矿业大学 安全工程学院，江苏 徐州  221116

摘要: 目的 针对气煤叠置区天然气管道引发的瓦斯泄露、环境污染等事故频发问题，以大牛地气田S矿、X2矿、H矿和M矿4个典型叠置矿区内管道为例，对该区域内天然气管道进行风险评价。  方法 通过分析管道破损原因，构建由管道因素、环境因素和人员管理因素3个一级指标，管道下沉量、管道腐蚀等17个因素为二级指标组成的风险评价体系；其次，基于可拓理论构建风险评价模型，利用BP神经网络计算指标权重，并分别基于距离关联度函数和相似度函数计算指标关联度，进而对管道进行风险量化分析。  结果 结果显示，基于距离关联度函数计算4个区域管道的风险关联度分别为0.002，0.034，0.053，0.019，基于相似度函数计算对应区域的风险相似度分别为0.054，0.510，0.560，0.500，4个区域对应风险等级分别为较大风险、一般风险、一般风险、低风险。  结论 基于距离关联度函数和基于相似度函数2种计算方法在可拓评价模型中均适用，评价结果基本一致，共同验证了评价结果的可靠性。

关键词:可拓理论;BP神经网络;气煤叠置区;天然气管道;风险评价

doi:10.16186/j.cnki.1673-9787.2024110030

基金项目:国家自然科学基金资助项目（52174109）；河南省重点研发专项项目（241111320800）；河南省自然科学基金资助项目（222300420170）；河南理工大学创新型科研团队项目（T2022-2）

收稿日期:2025/03/15

修回日期:2025/06/10

出版日期:2026/01/28

Study on safety risk assessment of natural gas pipelines in gas-coal overlap zones based on extension theory

Yue Shuaishuai1, Fan Wenhuan1, Wang Wen1,2, Guo Runsheng3, Lu Xiaowei4

1.School of Energy Science Engineering， Henan Polytechnic University，Jiaozuo  454003， Henan， China；2.Collaborative Innovation Center of Coal Work Safety and Clean High Efficiency Utilization， Jiaozuo  454003， Henan， China；3.North China Oil and Gas Branch，China Petrochemical Co.， Ltd.， Zhengzhou  450000， Henan， China；4.School of Safety Engineering， China University of Mining and Technology， Xuzhou  221116， Jiangsu， China

Abstract: Objectives In response to the frequent accidents such as gas leakage and environmental pollution caused by natural gas pipelines in gas-coal overlap zones， this study takes pipelines in four typical overlapping mining areas of S Mine， X2 Mine， H Mine， and M Mine in the Daniudi Gas Field as examples to conduct a risk assessment of natural gas pipelines in the regions.  Methods By analyzing the causes of pipeline damage， a risk assessment system was constructed， consisting of three first-level indicators （pipeline factors， environmental factors， and personnel management factors） and seventeen second-level indicators （such as pipeline subsidence and pipeline corrosion）. Subsequently， a risk assessment model was built based on extension theory. The BP neural network was employed to calculate the weights of the indicators. The correlation degrees of the indicators were then calculated using distance correlation functions and similarity functions， respectively， followed by a quantitative risk analysis of the pipelines. Results The results show that the risk correlation degrees for pipelines in the four areas， calculated based on the distance correlation function， were 0.002， 0.034， 0.053， and 0.019， respectively. The corresponding risk similarities calculated based on the similarity function were 0.054， 0.510， 0.560， and 0.500， respectively. The corresponding risk levels for the four areas were "high risk，" "moderate risk，" "moderate risk，" and "low risk." Conclusions Both calculation methods based on the distance correlation function and the similarity function are applicable in the extension evaluation model. The assessment results are largely consistent， jointly verifying the reliability of the evaluation outcomes.

Key words :extension theory; BP neural network; gas-coal overlap zone; natural gas pipeline; risk assessment