Prediction Of Rock Deformation And Analysis Of Rock Stability Of Tunnel Based On The Neural Network

Tunnel belongs to a special structural system. From rock mechanics point of view, it is in interaction with the surrounding structures of the system; from a geomechanical point of view, it is in the ever-changing geology of the project unit [1]. In such a rock and geological bodies, the tunnel and build the process of the mechanical properties of rock not only by conditions of rock formation and geological conditions, also affected by the tunnel excavation methods, support type, support the opportunity to support parameters, such as impact, which seek to correctly reflect the physical status of rock mechanical model is very difficult [2]. While in the tunnel design and construction of major reference works and mathematical analogy preliminary analysis, the construction process through real-time monitoring of rock on the monitoring data analysis and synthesis of judge, and further improve the design and construction to take corresponding countermeasures. Of these, monitor and control the work of measuring the results of analysis as a measure of the design, construction is reasonable an important job.This paper based on Bai ma tunnel of Yu Xiang highway project and ragard the highway tunnel surrounding rock displacement of deformation as the main line and systematic study surrounding rock stability and reliabilityFirst, according to the Bai ma Tunnel construction site monitoring monitoring data for the stability of surrounding rock deformation characteristics and to sum up the surrounding rock deformation of the influencing factors for a preliminary study, for the main engineering geological problems were evaluated.Second, this paper studies the existing NATM construction monitoring of the content of information on the construction site measurement data regression analysis, through analysis of measured data, will be in a better rock dynamics and supporting structure of the job status, and foreseeable accidents and dangerous situations in order to take timely measures, the construction of safety and to ensure the quality of the project plays an important role.Finally, through the white tunnel displacement measurement, this paper use neural network theory of BP neural network, Elman recurrent neural network of the tunnel displacement predicted. At the same time, the paper sets up a combination of prediction models, and analysis the prediction of model of three types then concluded that: Based on the theory of neural networks and a combination of gray prediction model predicte minimum error, more near the tunnel displacement laws, this shows that the model prediction for the displacement of the tunnel has a good adaptability.