A rapid dynamic classification prediction method for surrounding rock during tunnel construction

Abstract

A method for fast dynamic classification and prediction of surround rock during tunnel construction includes such step as constructing a mathematical model of the least square vector machine and a bacterial foraging algorithm, setting initial variables, obtaining measurement parameters, construct learning samples, and calculating BQ true value BQ0; obtaining the BQ prediction value BQ1 by solvingthe least square vector machine with the mathematical model of bacterial foraging algorithm; when BQ1 and BQ0 are close enough to each other, obtaining the optimal regular parameter variables and kernel parameter variables; establishing the nonlinear mapping relationship between measured parameters and BQ0 by the optimal canonical parameter variable and kernel parameter variable, and classifying the corresponding surrounding rock grades based on BQ1 referring to the national standard. In combination with the bacterial foraging algorithm and the nonlinear mapping function of the least square vector machine to classify the surrounding rock of tunnel, the method is robust, exhibits global optimization, and can accurately evaluate or predict the grade of surrounding rock.

Description

Technical field [0001] The invention relates to the technical field of surrounding rock safety evaluation, in particular to a fast and dynamic classification prediction method for surrounding rock during tunnel construction. Background technique [0002] The level of surrounding rock contains various information such as rock stability, mechanical properties, permeability, etc. During the construction of highway tunnels, the determination of the level of surrounding rock is an important basis for further construction and support plan selection. At present, the rock mass classification in the construction design process is mostly based on the classification results in the initial geological survey report of the project. At the same time, during the excavation process, the exposed face is evaluated and analyzed based on the experience and the national standard BQ method and other evaluation models. One-step construction plan and support means. Due to the complexity and nonlinearity...

AI Technical Summary

Problems solved by technology

Due to the complexity and non-linear characteristics of geotechnical engineering, there is often a certain difference between the revealed results of the surrounding rock grade and the initial geological survey and design results; the judgment results obtained by the conventional BQ method often have errors and poor robustness; Analysis and judgment require high experience and quality of surveyors and there are certain subjective uncertainties

Images