The Systemic Landslide Risk Analysis And Control In The Construction Of Highway Tunnels

With the rapid development of national economy in China, the highwaytransportation network becomes more integrate gradually. For the large numbers ofour country’s land area foothills distribution, the large number of tunnel engineeringsin highway construction began to appear quickly under the premise of economicrequirement and environmental protection. The construction safety accident frequencyappeared to be rising because of the relative covert and complex geologicalenvironment. The tunnel collapse tunnel in the construction process is the mosttypical disaster phenomenon and the cause is relatively complex. It is also consideredas one of the important problems by the researchers. It is carried out by the followingfour aspects of the research work in this article to solve the existing problems of riskanalysis method.First of all, the present situation and problems of landslide risk analysis duringthe tunnel engineering construction period is summarized detailedly, then it inducesthe applicable scope and limitation in the existing risk assessment method. Combinedwith the characteristics of the highway tunnel, it put forward the tunnel collapse riskanalysis and control process which covers each step briefly. Combined withsurrounding rock deformation theory, detailed analysis of the highway tunnellandslide mechanism is used to improve the flow chart of the theory basis.Secondly, the hole, cave and tunnel face collapse fault tree analysis diagramduring the road tunnel construction are prepared based on the participation inengineering practice.And it is screening out the critical collapse to find the integrationand sorting using the orthogonal experimental design method. On this basis more than117tunnel’s collapse information are collected, and the collapse factor classificationis clear combined with the appropriate statistical analysis and numerical simulationresults.Then the introduction of rough set theory can help us to calculate the weight ofthe objective set, and to avoid the lost phenomenon of the smaller weight factors.With the117detailed tunnel collapse accident statistics in the past15years, we canidentify the six major landslides risk factors including geological factors, design andconstruction quality and so on.By the application of rough set theory in the decisiontable method, we can also calculate the factors of the objective weight and use thefuzzy theory to focus on the evaluation factors as normal membership function approach. Then we can establish tunnel construction period collapse assessmentmethod based on the rough set theory and theory of barycenter of fuzzy set. Finally,using the assessment methods in the engineering project, we find this method canverify the applicability and it’s more accurate compared with the traditional fuzzycomprehensive evaluation and practical engineering analysis method.After it, by collecting117tunnels’ collapse statistics based on the introduction ofthe divide and conquer strategy in the rough set theory, we can screen algorithmsRSDIDA data and fill the incomplete collapse of the data set. We can also construct aneural network model for the prediction of tunnel construction collapse based on thegeneralized regression neural network theory, and use the Matlab Neural NetworkToolbox to find the best smooth factor by cross-validation and circuit training. Thenwe predict the landslides amount and verify its accuracy and applicability in ErZhuangke Tunnel’s risk analysis. Comparing the calculation results of the thegeneralized regression neural network collapse prediction method with the BP neuralnetwork algorithm, we find this comprehensive evaluation method with the tunnelcollapse risk assessment model shows more reliability and objectivity during theconstruction period.Finally, based on the principle of stability of surrounding rock, the ANSYSsimulation is used to the different conditions in certain projects with the deformationand displacement monitoring information, in order to achieve the optimization of theconstruction method and the proposed road tunnel collapse risk control system. Acomplete set of the tunnel construction of landslide risk analysis and control system isconstructed which contained the landslide risk fuzzy rating method, landslideprediction method and risk control system.The completion of this thesis is based on National natural scientific researchproject (No.51078136) and Hunan natural scientific research project (No.09JJ3113).