Abstract:

The quantitative theory Ⅲ is used to quantitatively analyze the deformation influencing factors and coupling of smallspacing tunnels; and a prediction model for tunnel deformation is established by using the limit learning machine to realize quantitative evaluation of influencing factors for smallspacing tunnel deformation, guarantee the effective analysis of tunnel deformation laws and verify the accuracy of the analytical results. The analytical results show that: (1) The dominant influencing factors for tunnel deformation include surrounding rock volume weight, deformation modulus, Poisson ratio, internal friction angle and cohesion; the important influencing factors include surrounding rock dilatancy angle, tunnel burial depth, shoctrete thickness, anchor bolt length and tunnel spacing; and the general influencing factor includes lateral pressure coefficient. (2) There is a coupling among every influencing factor; the coupling degree among the samples 4, 6, 16 and 18 is high; and the coupling degree among other samples is low. (3) The extreme learning machine prediction results indicate that the general influencing factor will weaken the correlation value between influencing factors and tunnel deformation, which is unfavorable for deformation law analysis; the results coincide with analytical results of deformation influencing factors; the extreme learning machine has a good deformation prediction results for each monitoring item, which verify the feasibility and applicability of the method for tunnel deformation prediction.

Key words: smallspacing tunnel, quantitative theory Ⅲ, limit learning machine, tunnel deformation, deformation prediction