摘要
为揭示安全隐患时空分布特征,挖掘某大型水电工程施工期间32060条安全隐患数据,采用基于词频-类别关联因素(TF-CRF)算法的短语提取技术,建立隐患类型和隐患部位的词云图,反映隐患关键信息的词频分布特征;绘制安全隐患关联桑基图,反映隐患特征、隐患部位、隐患类型之间的信息流动特征;构建安全隐患的时序复杂网络,分析不同隐患类型的时间特征及其动态联系。研究结果表明:施工排架、钢模台车、排架搭设等隐患在地下洞室、洞内、平台等部位是排查的重点;从桑基图可得,与钢模台车和施工排架相关的信息流最密集,与它们有关的隐患问题最多;由网络图可得,触电安全隐患在11月和12月发生最频繁,1月和2月发生次数最少;将3种隐患类型整合到同一网络进行验证,证明此方法可行,同时,发现与文明施工有关的安全隐患数量最多,这些应引起管理者高度重视。
In order to reveal the temporal and spatial distribution characteristics of safety hazards,the data of 32060 safety hazards during the construction of a large-scale hydropower project were mined.The phrase extraction technology based on the TF-CRF algorithm was used to establish a word cloud map of the types and parts of hazards,reflecting the word frequency distribution feature of key information about safety hazards.The Sankey diagram was produced to represent the information flow features among characteristics,locations and types of safety hazards.The sequential network of near-misses was constructed to analyze the time characteristics and dynamic relationships of different types of safety hazards.The results show that safety hazards such as the construction of the bent frame,steel formwork trolley,and bent frame erection frequently occur in underground caverns,caves,platforms and other parts,which are the focus of safety management.From the Sankey diagram,the information flow related to the steel formwork trolley and the bent frame is the densest,and the safety hazards related to them are the most.From the network diagram,Electric shock safety hazards occur most frequently in November and December,and the number of occurrences is the least in January and February.Integrating three types of safety hazards into the same network for verification shows that this method is feasible and also finds the highest number of safety hazards related to civilized construction.These should be attached with great importance by the managers.
作者
陈述
刘文濯
邵波
习俊博
王建平
CHEN Shu;LIU Wenzhuo;SHAO Bo;XI Junbo;WANG Jianping(Hubei Key Laboratory of Construction and Management in Hydropower Engineering,China Three Gorges University,Yichang Hubei 443002,China;College of Economic and Management,China Three Gorges University,Yichang Hubei 443002,China)
出处
《中国安全科学学报》
CAS
CSCD
北大核心
2022年第11期90-96,共7页
China Safety Science Journal
基金
国家自然科学基金资助(52079073)。
关键词
水电工程施工
安全隐患
时空分布
桑基图
时序网络
词频-类别关联因素(TF-CRF)算法
hydroelectric engineering construction
safety hazards
spatial and temporal distribution
Sankey diagram
sequential network
term frequency-category relevance factors(TF-CRF)algorithm