Research On Deformation Monitoring Prediction And Construction Dynamic Risk Assessment Of Deep Foundation Pit Of Subway

With the rapid development of China’s economy and the improvement of urbanization level;Currently,the function of urban ground transportation almost reaches the upper limit,so the most efficient solution to solve the space congestion is to develop urban railway transportation and to rationally develop underground space.Among them,deep foundation pit engineering as an important link in the construction of Metro station,is a complex and risky project due to the influence of many factors during the construction process.Therefore,through the monitoring work and correlation analysis of deep foundation pit,it can effectively control the safety and stability of foundation pit,prevent the occurrence of safety accidents,and provide guidance for the design and construction of deep Foundation pit engineering.Based on the deep foundation pit engineering of a subway station on Xiamen Metro Line No.3,considering about on-site monitoring data,and according to the research on the deformation laws,deformation prediction and risk assessment of deep foundation pit monitoring items.The specific works and conclusions of the thesis are as follows:1.This thesis expounds the monitoring scheme of deep Foundation pit in Metro station,selects the representative monitoring point in the important monitoring project for typical analysis,and draws the deformation curve of time-monitoring data through the Origin software,to analyzes the deformation law existing in it,through the information feedback and guidance construction,for the future with similar projects in the region to provide empirical reference.2.This thesis illustrates the principle,flow and characteristics of BP neural network and genetic algorithm,realizes BP neural network model and GA-BP neural network model through MATLAB programming,and applies it to the prediction of vertical displacement on the surface of deep Foundation pit.Comparing the forecast results with the field monitoring results,the average relative error predicted by the two models is 2.59%and 2.2%respectively,which indicates that the deformation prediction accuracy is high,which can be used to guide the construction to find the safety hidden trouble in the deep foundation pit in time,so as to take corresponding measures quickly.3.By comparing the risk assessment methods of foundation pit,such as analytic hierarchy process,expert evaluation method,Delphi method and Risk matrix method,this thesis determines the risk assessment method based on on-site monitoring data,combined with AHP and expert evaluation method,so as to establish the risk assessment system and realize the risk assessment of real-time dynamic of foundation pit by MATLAB programming.The evaluation results show that the comprehensive risk level of foundation pit is increased from level four to level three and remains unchanged,and different countermeasures can be taken according to different foundation pit states to ensure the safety and stability of Metro Deep Foundation pit during construction period.