Research On Safety Risk Management Of Adjacent Bridge Deformation Induced By Tunnelling

Metro construction is always located below the main road of the city, and it willinevitably arise the situation that metro construction goes through the existing bridge.Ground deformation caused by metro construction will have a huge impact on thestructural safety of adjacent bridge. Therefore, studying the effects of metro constructionon adjacent bridge safety has great theoretical significance and engineering applicationvalue. In this thesis, the Rough Sets theory is used for knowledge acquisition ofengineering test data. Based on the results of knowledge acquisition, a Bayesian Networkis constructed to assess the grade of safety risk of adjacent bridge caused by metroconstruction and analyze the key factors of the risk in construction process, which helps toprovide decision supports for safety management of metro construction.The thesis begins with a literature review of relevant studies on safety riskmanagement of adjacent bridge deformation caused by metro construction. Combinedwith the practice experience, main factors of adjacent bridge safety risk affected by metroconstruction is recognized, which covers four aspects such as the tunnel conditions, soilconditions, bridge conditions, construction methods and management level. Then RoughSets is used to find out the key risk factors of adjacent bridge safety affected by metroconstruction as well as the relationship between these factors. According to the results ofRough Sets knowledge acquisition, the Bayesian Network model of adjacent bridge safetyrisk in metro constructions is constructed. The results of forward reasoning of the testsamples validate the effectiveness of the Bayesian Network model, which means it couldbe used for practical application. Finally, taking a metro project as an example, anassessment of safety risk of adjacent bridges is carried out by applying Bayesian Networkreasoning. The key risk factors are achieved with using sensitivity analysis, and thecorresponding risk control measures are put forward. In this thesis, safety risk of adjacent bridge induced by metro construction was studied by combining Rough Sets knowledgeacquisition and Bayesian Network reasoning, providing a new idea for adjacent bridgesafety risk management.