| Along with the increasing speed of urbanization,it brings along the expansion of population scale,causing urban transportation to increase rapidly,resulting in a sharp increase in urban traffic pressure and other problems,the metro has the advantages of speed,efficiency and environmental protection,making the construction scale of metro projects gradually expand,and in the "14th Five-Year Plan" period has become an important means of building a strong transportation country in China.As a core component of metro projects,metro stations are characterized by complexity,environmental specificity and multi-technology intersection,which lead to frequent safety accidents during the construction stage.Therefore,how to improve the means of metro station construction safety risk management has become an urgent problem,for the existing metro station safety risk management there is insufficient use of safety historical data,safety risk control is mostly based on personal experience and other problems,this thesis from the metro station construction safety risk prediction and control and other aspects,to carry out systematic research,in order to help improve the level of safety risk to provide theoretical support,the main work of this thesis is as follows:Firstly,this thesis widely collects the textual data of safety accidents in metro stations,based on which,using the grounded theory,determines the influencing factors of metro station construction safety,and uses the questionnaire method to verify the usability and validity of the influencing factors through reliability analysis and validity test,aiming to use the historical data information scientifically,objectively and effectively,and to identify and analyze the influencing factors of metro station construction safety from a new perspective.Secondly,starting from a large number of historical metro station construction safety accident cases,the metro station safety risk prediction model is constructed,and advanced machine learning algorithms such as GWO-SVM and Fuzzy c-means are used to predict the safety risk level of the proposed station,and according to the applicability of the prediction methods,a suitable prediction method is selected under the condition of different data sample sizes,and under the condition of sufficient sample size,the GWO-SVM algorithm is used to predict the safety risk level;under the condition of insufficient sample size,the Fuzzy c-means clustering algorithm is used to predict.Finally,the safety risk control system of metro stations is constructed,and Near-miss events with a high number of occurrences and a small degree of loss are selected as the research objects,and statistical analysis and PDCA cycle are used to study the near-miss events occurring in the construction phase of metro stations,so as to realize the dynamic control of safety risks in metro station construction.On this basis,a safety risk management system is established and the process of safety near-miss management is designed to avoid the evolution of near-miss events into safety accidents.And the BIM 3D visualization model is used to assist safety risk management,which improves the level and efficiency of safety risk management. |
