| The surrounding rock classification is the basis of highway tunnel construction, its accurate division is of important significance to optimum deign for tunnel structure and construction safety. However, during the construction of highway tunnel in karst areas, it is very difficult to accurate determination of the surrounding rock grade with the complexity of engineering geological conditions and the influence of karst development. Meanwhile, it is lack of classification method for this special rock at present. All of above had made a serious barrier to the development of highway tunnel construction in karst.Based on the popular classification index system of surrounding rock at home and abroad, combined with the special nature of karst rock, this paper proposed a surrounding rock classification index system which is suitable for the highway tunnel construction in karst areas. This indicator system included5indexes of rock strength, rock mass integrity, groundwater state, strike-dip of structural surfaces and karst state.This paper also proposed a method to get each index. Then, relied on the project construction of Yaozhai tunnel, Tianshengqiao tunnel and Guangshang2tunnel, it built a neural network expert knowledge base with the established index system. Later, according to the genetic neural network mathematical theory, used the numerical calculation software MATLAB to construct the karst rock classification expert system model. Follow, using mixed programming technology of MATLAB and C++, as the hierarchical model to be core, based on the development platform of Visual C++6.0, completed the development of surrounding rock classification expert system for highway tunnel in karst areas.To test and verify the reliability of this expert system, used it in the Jingna highway, and compared with the actual results. Practice shows that the system’s accuracy has reached80.89%, which can basically meet the requirements of tunnel construction. |
PDF Full Text Request