Intelligent And Non-linear Analysis Of The Stability Of Deep-buried Tunnel

Alongside implementation of the strategy of West Development, China meets the large-scale constructions of infrastructures and exploitations of energy, such as South-to-North Water Transfer Project, Electronic Power Transmission from West to East, Qinghai-Tibet Railway, Transporting Natural Gas from West to East and West Traffic Infrastructure Construction. Along with the constructions of these important projects, long deep-buried tunnels and large underground engineering are inevitable. In order to reveal the mechanism of the failure of the deep-buried tunnel's surrounding rocks and adapt to the development of the tunnel's designing and constructing, it is necessary to research the stability of deep-buried tunnel's surrounding rocks systematically.The paper was supported by the"Research on the stability and reliability of tunnel and underground space structure"(50334060), a key project of National Nature Science Foundation of China. The main actual engineering of the paper is Tongyu Tunnel in Chongqing. The paper uses artificial intelligence and catastrophic theory to analyze the stability of the deep-buried tunnel's surrounding rock. The main conclusions of the paper are showed as following:(1) On the base of discussing the displacement criterion of tunnel's stability and characteristics of time series of tunnel's displacement, the paper analyzed the principle of Gray Model, Chaotic Time Series Model and mixed optimizing model for prediction of the tunnel's displacement. The predicting results of these displacement-predicting models agree well with the measured results. The research has good academic and practical significance.(2) The mixed process of training and evolution of BP neural network are realized by decimal genetic algorithm. The evolution of BP neural network includes the structure of hidden layer and learning parameters. Visual C++6.0 is used to realize the actual program.(3) The paper proposed a displacement-predicting model of deep-buried tunnel based on evolutionary neural network. And the new model was applied in Tongyu Tunnel. The predicting results show that the model can get high precision efficiently. The efficiency of the model increases 3~10 times than BP neural network.(4) Intelligent-identifying model for classification of deep-buried tunnel's surrounding rocks based on evolutionary neural network was proposed. The validity of the model is good. It shows that intelligent-identifying models for classification of...