Study On The Prediction And Visualization Control Of Safety Risks In Deep Foundation Pit Constructions

In recent years,the development of metro constructions in China has come into the golden period.With the rapid development of the metro constructions,the safety accident during the construction process has occurred frequently,which made a serious threat to the surrounding environment and the residents' property safety.The reason of frequent accidents is not only the characteristics of the foundation pit-“big,tight,deep”,but also the shortcomings of risk management,such as inaccurate risk prediction,inaccurate risk laws,not targeted risk control and low risk management level.Therefore,this paper studies the risk prediction,the occurrence law and the visualization method of safety risks in metro constructions,the systematic risk management method was proposed.The proposed method was verified by taking the Wuhan Metro station as an example.The results show that the method can predict the risk states accurately,implement analysis of risk causes and visual control.Firstly,aming to the current situation of risk management,this paper focused on the development of risk management methods and evaluation methods at home and abroad(mainly empirical,numerical simulation methods,machine learning methods).The advantages and disadvantages of the three types of methods were discussed emphatically.It was determined that Support Vector Machine(SVM)was used for risk prediction,Bayesian Networks(BN)was used for risk diagnosis,Building Information Model(BIM)was used for risk visualization management and control.Secondly,the basic framework and process of the risk management were constructed based on the relevant theories.The steps of risk prediction based on SVM,the analysis process of risk diagnosis based on BN,and the flow of risk visible control based on BIM were given in detail.Through the combination of the three methods,feed-forward,concurrent and back-forward risk control could be achieved.Finally,taking the Wuhan Metro station as an example,on the basis of risk identification,SVM was used to predict the risk levels,which could effectively solve the difficulty in data collection of foundation pit engineering.Then,according to the results of risk prediction,the temporal and spatial evolution laws of the overall construction risk in the metro station were obatained.At the same time,based on the prediction results,BN was used to analyze the reasons of diaphram wall sway.The key factors leading to the diaphram wall sway are determined,so the risk control could be performed targetly.At last,BIM technology was introduced for risk control in foundation pit.The risk states can be displayed in real time,the risk response reports were generated automatically,and the key construction process can be simulated in advance.Through the actual engineering case,it was proved that the method can effectively predict the risks in metro constructions and provide effective guarantee for the safety control in construction process.