Font Size: a A A

Study On Safety Risk Pre-Control Of Underground Engineering Crossing Existing Subway Construction Based On GA-BP

Posted on:2021-01-01Degree:MasterType:Thesis
Country:ChinaCandidate:J MengFull Text:PDF
GTID:2381330611488826Subject:Civil engineering construction and management
Abstract/Summary:PDF Full Text Request
With the increasing density of urban underground transportation network and the increasing scarcity of construction land,it is more and more common for underground projects to pass through existing subway tunnels.In the process of such engineering construction,the surrounding soil will be disturbed,which will lead to the deformation of the subway tunnel.Once the deformation is too large,it will not only lead to the rupture of pipe segments and tunnel leakage,but also bring great safety risks to the operation of trains.Therefore,in addition to taking some passive control measures in advance,it is also important to timely know the deformation and development of the subway tunnel,so as to gain valuable time for the subsequent active control measures.For this reason,the paper analyzes the factors that affect the deformation of existing subway tunnels during the construction of underground engineering,establishes a corresponding forecast index system,and builds a safety risk prediction model for underground engineering crossing existing subway construction.It was validated and analyzed by case engineering,and the validity of the prediction model was confirmed.Then,field tests were conducted to study control technologies such as grouting and back pressure to provide a basis for formulating corresponding pre-control measures.First of all,according to the differences between the sources of influence factors,from the aspects of underground engineering excavation,hydrogeology conditions,and the status of the existing subway tunnel,the factors affecting the deformation of the subway tunnel during the construction of the underground engineering Crossing the existing subway are analyzed Impact,and then through quantitative research and classification analysis of each influencing factor,a construction safety risk prediction index system for underground engineering crossing existing subway tunnels is established.Secondly,based on the characteristics of underground engineering crossing existing subway tunnel construction,the genetic algorithm was used to optimize the BP neural network to establish a safety risk prediction model for underground engineering crossing existing subway tunnel construction,and real-time prediction was achieved through the "metabolism" method.The empirical analysis shows that: first,the prediction model has good prediction accuracy;second,the genetic algorithm has played a good role in optimizing the prediction model;third,the prediction model has good generalization ability and practicability;Fourth,the model still has a good prediction ability after extending the prediction time.Finally,according to the practical project and field test,the influence of the construction of direct excavation-shoring and cover-digging on the subway tunnel is studied,as well as the effects of passive control technologies such as "partition skipping" construction and active control technologies such as "grouting backpressure".And on the basis of the above research to develop safety risk pre-control measures.The results show that: First,the effect of direct excavation-shoring construction on tunnels is significantly greater than that of cover-digging construction.The direct excavation-shoring construction uses the “Zoning and jumping” excavation method to effectively utilize the “space-time effect” in the excavation process Principle to reduce the impact of the excavation on the tunnel.The cover-digging construction uses the ground connection wall as the underground side wall,which can effectively reduce the lateral movement of the ground connection wall.Secondly,timely active grouting can effectively reduce the convergence deformation of the tunnel,and can significantly suppress the subsequent development of convergence deformation.Thirdly,timely and active back pressure above the tunnel can effectively reduce the tunnel uplift rate,but it will cause the tunnel's convergent deformation to continue to increase.This back pressure is only applicable to sections with excessive uplift deformation and small convergence deformation.
Keywords/Search Tags:underground engineering, Existing subway tunnel, BP neural network, Genetic algorithm, Risk prediction, Risk control
PDF Full Text Request
Related items