| 118 | 0 | 189 |
| 下载次数 | 被引频次 | 阅读次数 |
地铁施工具有高度动态性与复杂性，坍塌事故风险突出。为了提升地铁施工安全管理水平并降低坍塌事故的发生概率，提出一种基于数据挖掘与贝叶斯网络相结合的事故分析方法。首先，以2001年1月—2025年9月120起国内地铁施工坍塌事故为样本，从人、设备设施、管理、环境识别了27个致因因素；其次，采用SPSS Modeler基于Apriori挖掘致因37条关联规则，并据此构建坍塌事故贝叶斯网络模型；最后，借助Netica开展概率评估、敏感性分析与逆向推理，并以变异比量化不同事故等级情景下的致因影响强度。结果显示：数据挖掘能够有效揭示地铁施工坍塌事故致因的关联规律；不同致因因素对坍塌事故发生概率与等级演化的边际作用与归因强度存在显著差异；坍塌事故对不良地质条件、违规作业与地下水侵蚀最为敏感；研究结果可为制定针对性的坍塌事故防控策略，降低坍塌事故的发生概率及其潜在危害提供借鉴。
Abstract：Subway construction is characterized by high dynamism, technical complexity, and significant environmental coupling, making collapse accidents one of the most critical safety challenges in underground engineering. To enhance safety management in subway construction and reduce the likelihood of collapse events, this study proposes an integrated accident analysis approach that combines data mining with Bayesian Network(BN) modeling. We collected a dataset of 120 collapse cases that occurred in China between January 2001 and September 2025. Within a four-dimensional framework—comprising human factors, equipment/material, management, and environment—we identified 27 causative factors. Using SPSS Modeler and the Apriori algorithm, we extracted 37 association rules to reveal the interdependencies among these factors, which subsequently informed the construction of a Bayesian network model for subway collapse accidents. The Netica platform was then utilized to conduct probability evaluations, sensitivity analyses, and reverse inference. Additionally, the Ratio of Variation(ROV) was applied to quantify the influence strength of different factors under various levels of accident severity. The findings demonstrate that data mining effectively uncovers the associative mechanisms underlying collapse causation. Key root causes of mild collapses were identified as improper construction methods, inadequate safety training, and managerial negligence. In contrast, excessive soil disturbance, inaccurate monitoring, and flawed design schemes significantly increased the likelihood of moderate and severe collapses. Under the assumption of collapse occurrence, the posterior probabilities of most factors progressively rose from minor to major levels. Specifically, factors such as unfavorable geological conditions, insufficient supervision, unsafe behaviors, and inadequate survey accuracy exhibited large absolute ROV values, confirming their dominant roles in collapse causation. The system was found to be most sensitive to adverse geology, unsafe operations, and groundwater erosion. These results provide valuable insights for formulating targeted prevention and control strategies aimed at reducing both the frequency and severity of collapse accidents in subway construction.
[1] 李政道，郭振超，苏栋，等.利益相关者视角下地铁施工及运营风险管理研究[J].安全与环境学报，2023,23(8):2819-2829.Li Zhengdao,Guo Zhenchao,Su Dong,et al.Research on subway construction and operation risk management from a stakeholder perspective[J].Journal of Safety and Environment,2023,23(8):2819-2829.
[2] 贺银娟.中部城市地铁发展对经济发展的影响研究[J].城市轨道交通研究，2023,26(9):282-284.He Yinjuan.A study on the impact of subway development on economic development in central cities[J].Urban Mass Transit,2023,26(9):282-284.
[3] 吴波，蔡琦，刘聪，等.城市地铁车站施工安全风险多尺度评估模型及应用[J].安全与环境学报，2023,23(3):633-641.Wu Bo,Cai Qi,Liu Cong,et al.Multi-scale evaluation model and application of safety risk in urban subway station construction[J].Journal of Safety and Environment,2023,23(3):633-641.
[4] 申建红，刘树鹏.基于TAN网络的地铁区间与车站施工事故致因分析[J].隧道建设(中英文),2023,43(1):27-35.Shen Jianhong,Liu Shupeng.Analysis of causes related to construction accidents in metro tunnel and station based on tree augmented naive Bayesian[J].Tunnel Construction,2023,43(1):27-35.
[5] 王乾坤，朱科，郭佩文.DEMATEL和模糊认知图在地铁深基坑施工安全风险动态评估中的应用[J].安全与环境学报，2024,24(11):4143-4153.Wang Qiankun,Zhu Ke,Guo Peiwen.Application of DEMATEL and fuzzy cognitive map in dynamic assessment of safety risk in subway deep foundation pit construction[J].Journal of Safety and Environment,2024,24(11):4143-4153.
[6] 陈绍清，熊思斯，何朝远，等.地铁深基坑坍塌事故安全风险分析[J].安全与环境学报，2020,20(1):52-58.Chen Shaoqing,Xiong Sisi,He Chaoyuan,et al.Safety risk analysis for the deep foundation pit collapse accidents in the subway construction system[J].Journal of Safety and Environment,2020,20(1):52-58.
[7] Zhang Ying,Gu Yingying,Lian Ningning,et al.Analysis of urban subway construction collapse disaster chain and research on chain breaking and disaster reduction[J].PLoS one,2025,20(3):e0318269.
[8] 吴贤国，冯宗宝，王洪涛，等.基于动态元网络的地铁施工安全事故致因分析[J].中国安全生产科学技术，2019,15(3):141-147.Wu Xianguo,Feng Zongbao,Wang Hongtao,et al.Analysis on causes of safety accidents in subway construction based on dynamic meta-network[J].Journal of Safety Science and Technology,2019,15(3):141-147.
[9] 郭文亚，周志鹏.地铁工程建设安全事故致因网络模型构建与解析[J].中国安全科学学报，2020,30(1):155-161.Guo Wenya,Zhou Zhipeng.Subway construction accident causation network modeling and analysis[J].China Safety Science Journal,2020,30(1):155-161.
[10] 黄雄，胡祖祥.地铁施工事故统计分析与研究[J].中国安全生产科学技术，2022,18(9):218-224.Huang Xiong,Hu Zuxiang.Statistical analysis and research of subway construction accidents[J].Journal of Safety Science and Technology,2022,18(9):218-224.
[11] 房明，张毅，谭玥.多因素耦合失效模式下地铁施工安全系统与管理模型[J].中国安全科学学报，2024,34(5):82-90.Fang Ming,Zhang Yi,Tan Yue.Safety system and management model of subway construction under multi-factor coupling failure mode[J].China Safety Science Journal,2024,34(5):82-90.
[12] 黄锐，谢朝武，黄松山，等.基于Apriori算法的中国游客滑雪损伤特征与成因机制分析[J].中国生态旅游，2023,13(5):853-869.Huang Rui,Xie Chaowu,Huang Songshan,et al.Analysis of the characteristics and causal mechanisms of skiing injuries among Chinese tourists based on Apriori algorithm[J].Journal of Chinese Ecotourism,2023,13(5):853-869.
[13] 卢有飞，冯国平，卢宾宾.基于关联规则挖掘技术的电网故障风险要素分析[J].武汉大学学报(工学版),2024,57(6):792-797.Lu Youfei,Feng Guoping,Lu Binbin.Analysis of risk factors of power grid faults based on association rule mining technology[J].Engineering Journal of Wuhan University,2024,57(6):792-797.
[14] Shao Lijia,Guo Shengyu,Dong Yimeng,et al.Cause analysis of construction collapse accidents using association rule mining[J].Engineering,Construction and Architectural Management,2023,30(9):4120-4142.
[15] Burzawa A,Bodet L,Dhemaied A,et al.Detecting mechanical property anomalies along railway earthworks by Bayesian appraisal of MASW data[J].Construction and Building Materials,2023,404:133224.
[16] Hu Zhuo,Dang Chao,Wang Lei,et al.Parallel Bayesian probabilistic integration for structural reliability analysis with small failure probabilities[J].Structural Safety,2024,106:102409.
[17] Le Yang.Research on data resource management of biomass energy engineering based on data mining[J].Energy Reports,2022,8:1482-1492.
[18] 高扬，曹媛.基于关联规则挖掘的直升机事故/事件分析[J].安全与环境学报，2020,20(3):849-856.Gao Yang,Cao Yuan.Analysis of the helicopter accidents based on the association rules mining[J].Journal of Safety and Environment,2020,20(3):849-856.
[19] 金辰浩，李铁，关磊，等.基于贝叶斯网络的储油罐区事故情景分析[J].中国安全生产科学技术，2024,20(8):120-128.Jin Chenhao,Li Tie,Guan Lei,et al.Scenario analysis of oil storage tank farm accidents based on Bayesian network[J].Journal of Safety Science and Technology,2024,20(8):120-128.
[20] 赵振武，贾朋霖.融合改进D-S理论与贝叶斯网络的机场旅客风险研究[J].安全与环境学报，2023,23(12):4425-4434.Zhao Zhenwu,Jia Penglin.Research on airport passenger risk by fusing improved D-S theory and Bayesian network[J].Journal of Safety and Environment,2023,23(12):4425-4434.
[21] 陈坤，陈序，魏鑫，等.基于贝叶斯网络的页岩气井喷风险分析[J].安全与环境学报，2019,19(6):1876-1883.Chen Kun,Chen Xu,Wei Xin,et al.Bayesian network-based risk analysis on the blowout of the shale gas wells[J].Journal of Safety and Environment,2019,19(6):1876-1883.
[22] Wei Xiang,Zhou Wenxing.Bayesian network model for predicting probability of third-party damage to underground pipelines and learning model parameters from incomplete datasets[J].Reliability Engineering & System Safety,2021,205:107262.
[23] 胡亚平，张认认，李新宏，等.基于贝叶斯参数学习的海底原油管道腐蚀状态评估[J].中国安全生产科学技术，2023,19(11):93-99.Hu Yaping,Zhang Renren,Li Xinhong,et al.Assessment on corrosion state of subsea crude oil pipeline based on Bayesian parameter learning[J].Journal of Safety Science and Technology,2023,19(11):93-99.
[24] 王亮.基于贝叶斯网络的铁路交通事故预警研究[J].中国安全科学学报，2022,32(增刊1):134-139.Wang Liang.Research on railway transportation accidents warning based on Bayesian network[J].China Safety Science Journal,2022,32(Suppl.1):134-139.
[25] Zarei E,Azadeh A,Khakzad N,et al.Dynamic safety assessment of natural gas stations using Bayesian network[J].Journal of Hazardous Materials,2017,321:830-840.
[26] 鲁义，伍江乐，邵淑珍，等.基于贝叶斯网络的危化品道路运输事故推理模型[J].中国安全科学学报，2022,32(3):174-182.Lu Yi,Wu Jiangle,Shao Shuzhen,et al.Prediction model for road transport accidents of hazardous chemicals based on Bayesian network[J].China Safety Science Journal,2022,32(3):174-182.
基本信息:
DOI：10.13637/j.issn.1009-6094.2025.1256
中图分类号:TP18;TP311.13;U231.3
引用信息:
[1]邓勇亮,周琦,刘泽东,等.基于数据挖掘和贝叶斯网络的地铁施工坍塌事故分析[J].安全与环境学报,2026,26(07):2650-2657.DOI:10.13637/j.issn.1009-6094.2025.1256.
基金信息:
国家自然科学基金项目(72201188); 苏州市科技计划(基础研究)项目(SJC2023002)
2026-07-15
2026-07-15