Study On The Initial Geo-Sress Of The GaLongla Tunnel Region

With the rapid development of the world economy,the underground project such as highway tunnel has developed quickly in the need of the traffic industry .At the same time, complicated geology condition also give great challenge for their development; especially the stability of hydraulic underground caverns will be directly affected the success of the project construction and operation benefits of the project, and the stability of the underground project closely related to initial geo-stress of the rock mass.Therefore, whether the use of initial stress field is reliable will directly affect the reliability and security of the design and construction of underground project.This paper is designed to get the reliable distribution of the initial geo-stress regions in the tunnel region, to guide the project design and construction. Taking the GaLongla tunnel region for example, the distribution of the initial geo-stress region in the tunnel region was got by use of some methods comprehensively such as linear regression analysis, finite element analysis, and neural network. According to some theoretical experience, I have done some discrimination in the region whether it is a high initial geo-stress region and forecast analysis to the formation of a rock burst. Major studys are as follows:Firstly, Studies and summarizes the cause and influencing factors of the initial geo-stress,basic distribution law of the initial geo-stress,calculation of the initial geo-stress and so on research whose situation in home and abroad presently. At the same time introduces the method of hydraulic fracturing which is used most today in the world in measureing the initial geo-stress.Secondly, this paper analyses the linear regression analysis method, analyses and compares the influence to the recurrence result of the different methods imitating of horizontal tectonic movement, and does a regression analysis of initial geo-stress in GaLongla tunnel region,brings forward six different imitating methods, and has carried on the analysis fairly to every method. At last finds out the method that is fit for this region best.Thirdly, because of the complex mapping relation between initial geo-stress and its influencing factors, and the strong in-out ability of ANN(Artificial Neural Network), establish the relationship between stress and its influencing factors by BP newal network, and evaluate its merits and drawbacks. In the linear regression method, the regression coefficients have no exact physics meaning and need hypotheses that rock mass is in elastic state. Meanwhile, in the ANN method, the training samples are difficult to generate. Combining the linear regression method with ANN method can solve these problems perfectly.Fourthly,base on the initial geo-stress of GaLongla tunnel region, analyses the effects on the stability of underground projects caused by the initial geo-stress of the rock mass and brings forward some measures in preventing and cureing geological disasters.Lastly, gives rise to the rock bursts easily in the high initial geo-stress regions, but is not all rocks burst give rise in the high initial geo-stress region, sometimes the low initial geo-stress region also have rock burst.