Research On Back Analysis Of Rock’s Parameter And Informative Construction In The Tunnel Engeneering

The research methods of traditional rock mechanics is the concern of deterministic system. However, application of pure theory and positive numerical prediction is hard work for the geotechnical engineering problems.Take advantage of monitering data to optimize the parameters is very important in geotechnical engineering.This paper having LvChun Ba railway tunnel as a background of engineering application, to use existing monitoring data as a basic information, combined with numerical analysis and calculation method of artificial intelligent to inverse analysis mechanical parameters of surrounding rock.Make regression analysis and curve-fitting on the observation date of typical cross section to find out that the change of surrounding rock mass deformation characteristics with time and space, predicted trend of surrounding rock deformation to optimize second lining time.The mechanical parameters of surrounding rock was designed by orthogonal and uniform experiment.Construct a3D finite element tunnel excavation model and combining with fractional steps excavation method for surrounding rock elastic-plastic analysis, multiple parameters sample will be put in the numerical model, figure out of the displacement of the key point which is used as BP neural sample.BP neural network easily get into the local minimum value and the training time is too long, it’s a shortcomings, take advantage of genetic algorithm global optimization ability to optimize BP neural network weights, initial matrix.Based on the comprehensive research on those mentioned above, GA-BP intelligent back analysis system applied in tunnel excavation engineering,combined with the surrounding rock mechanics parameters, the inversion optimization parameters generation into the finite element program, evaluation and analysis the stability of surrounding rock and supporting structure.