| With the continuous development and increasing scale of my country’s construction industry,due to the characteristics of complex construction procedures,personnel movement,and harsh environment,the safety production accident rate is still high.Therefore,it is of great theoretical value and practical significance to clarify the complicated interactions and correlations between the causes of construction safety production accidents,and to carry out early warning and classification of accidents to achieve ex-ante control.Based on the STAMP,this paper starts with a qualitative analysis of the causal factors of building accidents starting from the safety control structure of the construction project;The method of combining the gray DEMATEL and ISM is used to classify the complex logical associations between building accidents,and to build a multi-level hierarchical model;The BN is used for network learning and reasoning analysis to determine the early warning level of construction accidents,and to perform early warning analysis on key causal factors and maximum risk paths.The correlation analysis method and hierarchical early warning model were used to analyze the "11.24" collapse accident in Fengcheng,Jiangxi,which has reference value for the same type of accident.The main work of this article is as follows:(1)According to the STAMP model,the construction safety control structure of five control levels including government level,enterprise management level,intermediary level,basic level and physical level is constructed;nine kinds of responsibility subjects including material and equipment suppliers in physical layer,employees in basic level,construction enterprises in management layer,third-party intermediaries in intermediary layer and government agencies in government layer are analyzed and identified systematically;28 factors including the lack of material files,improper protection of materials,and poor maintenance of equipment are finally selected.(2)The gray DEMATEL/ISM quantitative correlation analysis model is used to clarify the interrelationships,causation paths and causal attributes,and establish the causes of 14-level hierarchical model of construction safety accidents including the causes of the neighbors,transitional causes,and essential causes.The causes of the neighbors are mostly caused by the physical layer and the grassroots,which are the direct causes of the accident;most of the transitional causes are caused by the management,and play a role in linking the interaction between the causes at the physical,grassroots and government levels;the essential cause belongs to the government level,which has a strong influence and control on the intermediary layer,the management layer,and the grassroots level.(3)The BN is used to carry out early warning analysis on the causes of construction safety accidents,and obtain the key causes and longest chain of accidents through network learning and inference analysis.The key causes include government agencies at the government level,construction units,supervision units,and construction units at the management level,the causes on the longest cause chain include the investigation unit and the design unit of the management,the third-party intermediary agency,and carry out early warning analysis from the perspective of these seven responsible entities and propose corresponding measures.(4)The above quantitative analysis method and hierarchical early warning model are used to analyze the "11.24" collapse accident in Fengcheng,Jiangxi.Based on the stamp model,the safety control structure of the third phase expansion project of Jiangxi Fengcheng power plant is constructed;The gray DEMATEL/ISM quantitative analysis method is used to build the 11 layer hierarchical model of collapse accident;The BN is used to carry on early warning analysis,find out the key cause factors and the maximum risk path of the accident,in order to provide decision-making basis for the construction project safety management. |
PDF Full Text Request