| 本文已被：浏览 1506次 下载 801次 |
码上扫一扫！ |
|
|
| 数据驱动的城镇燃气管道风险因素识别及关联分析 |
|
张晓雷1,郑春宏2,刘露3,葛彦泽3,徐小峰3,黄玉萍1,马俊4
|
|
(1.中国石油天然气股份有限公司天然气销售山东分公司,山东 济南 250013;2.青岛中石油昆仑胜利燃气有限公司,山东 青岛 266100;3.中国石油大学(华东) 经济管理学院,山东 青岛 266580;4.中国石油天然气股份有限公司天然气销售分公司 第二质量安全环保监督中心,北京 100034)
|
|
| 摘要: |
| 城镇燃气管网作为重要的满足群众生活需求的基础设施,其安全性对于保障广大人民群众的财产与生命安全具有重要意义。以810件国内城镇燃气管道泄漏、爆炸事故案例为样本,采用文本挖掘技术对样本案例进行分词处理,根据TF-IDF算法统计词频并确定导致燃气管道泄漏的关键风险因素；通过共现分析实现风险因素间关系的可视化,计算中心性指标,确定风险因素集合；基于Apriori算法揭示了燃气管道安全风险因素之间的关联规则。研究发现:在城镇燃气管道关键风险因素识别过程中,文本挖掘方法与传统方法分析结果基本一致；在次要因素分析中,管道设备老化、安全生产过程中的员工培训、安全保护措施、审批程序不完善等因素,相较以往研究都呈现出了与燃气事故更强的关联程度,这为管道安全管理提供了新的视角。 |
| 关键词: 城镇燃气管道 风险因素 文本挖掘 共现分析 关联规则挖掘 |
| DOI：10.13216/j.cnki.upcjess.2024.03.0002 |
| 分类号:TU996 |
| 基金项目:中石油昆仑燃气有限公司山东分公司科学研究与技术开发项目 |
|
| Data-Driven Urban Natural Gas Pipeline Risk Factor Identification and Correlation Analysis |
|
ZHANG Xiaolei1, ZHENG Chunhong2, LIU Lu3, GE Yanze3, XU Xiaofeng3, HUANG Yuping1, MA Jun4
|
|
(1.Shandong Branch of PetroChina Natural Gas Sales Co., Ltd., Jinan 250013, Shandong, China;2.Qingdao PetroChina Kunlun Shengli Gas Co., Ltd., Qingdao 266100, Shandong, China;3.School of Economics and Management, China University of Petroleum (East China), Qingdao 266580, Shandong, China;4.Second Quality Safety and Environmental Protection Supervision Center, PetroChina Natural Gas Sales Co., Ltd., Beijing 100034, China)
|
| Abstract: |
| As an important infrastructure to meet the needs of people 's life, the safety of the urban gas pipeline network is of great significance to protect the property and life safety of the general public. With 810 cases of domestic urban gas pipeline leakage and explosion accidents as samples, the text mining technology is used to process the sample cases, and the key risk factors leading to gas pipeline leakage are identified according to the TF-IDF algorithm with word frequency statistics; the visualization of the relationship between risk factors is realized through the co-occurrence analysis, and the centrality indicators are calculated to determine the set of risk factors; the association rules between the safety risk factors of gas pipelines are revealed based on the Apriori algorithm. Based on Apriori algorithm, the correlation rules between gas pipeline safety risk factors are revealed. It is found that in the process of identifying the key risk factors of urban gas pipelines, the text mining method is basically the same as the traditional method; in the analysis of the secondary factors, the aging of pipeline equipment, staff training, safety protection measures, and imperfect approval procedures in the process of safety production show a stronger correlation with gas accidents compared with the previous studies, which provides a new perspective for the management of pipeline safety. |
| Key words: urban gas pipeline risk factors text mining co-occurrence analysis association rule mining |
