Research On Safety Management Of Subway Construction For Accident Network Modeling

Subway and other public transportation tools have become the first choice for more and more people to travel.Major cities also take subway construction as an important development direction.However,in recent years,there have been serious subway construction accidents,often accompanied by personal injuries and equipment losses,which are urgent problems that need to be resolved.In the past,most accident analysis researches only focused on the causation factors of a single accident or a type of accident,and lacked research on the complex interactions between causation factors and different types of accidents during the accident.Therefore,it is great significance for improving the level of subway construction safety management to reveal the evolution between causation factors and accidents in the process of accident and avoid similar accidents.Firstly,this paper forms a set of standardized classification standards for the accident causation and accidents by integrating the provisions of national standards and the characteristics of subway accidents,as the basis of accident case analysis.Using the network theory to explore the mechanism of accident occurrence and internal cause mechanism,the causation-accident chain was coded based on the collected data of 486 subway construction accidents,and the coding effect of the model was verified by graph structure similarity measurement.Finally,a double-layer Subway Construction Accident Causation Network(SCACN)model with 56 causation nodes and 44 accident nodes were obtained.Secondly,the topological structure of the network model is analyzed macroscopically using the network topological feature quantity.Calculate and analyze the selected parameters such as degree distribution,cumulative degree distribution,average shortest path,and betweenness centrality,to characterize the scale-free network characteristics and small-world characteristics of subway construction accident causation network.Through random target attacks and specific target attacks,the cascading failure effect of nodes after attack is studied to determine the results of disturbances to the network.By projecting the double-layer network into a single-layer network,the relevant properties of the projection network are analyzed.It is found that eliminating or weakening the Hub node in the causation factors during the safety management of subway construction will minimize the occurrence of subsequent accidents and help improve the safety management level of subway construction.Finally,the Bayesian network is used to quantify the risk of subway construction accident causation network model.On the basis of constructing the SCACN model,the forward probability causality reasoning of the Bayesian network is used to predict the probability of the occurrence of the safety accident risk of subway engineering construction under different causal factors or native accidents;the cause mechanism of subway construction accident is analyzed by reverse diagnostic reasoning to determine the most likely causation factors or primary accidents.Quantitative research on the risks of the SCACN model from multiple different perspectives can help improve safety managers’ understanding of accidents and the evolution of risks,so as to do a good job of accident prevention and risk diagnosis.