Knowledge Mining of Classification of Railway Tunnel Surrounding Rock
Zhaoyi Xu · Journal of the China Railway Society · 2007
It is seen from various classification of surrounding rock at home and abroad that there is close relationship between discontinuities and classification of surrounding rock and different classification of surrounding rock possesses different characteristics of discontinuities.This paper puts forward the way of mining knowledge for surrounding rock classification through qualitatively analyzing the relationship between discontinuities and surrounding rock classification.According to the data mining principles,apparently unsuitable attributes of mechanics and lithology are firstly removed,and then,the needed attributes are selected and mined with information increasing.The concept of discontinuity weights and measures is presented to express incisions.The information increasing analysis indicates the roles of attributes of common discontinuities and faults differ to a great extent in surrounding rock classification and so different attributes are selected.The concept description and concept comparison description are expressed in forms to figure out the contribution magnitude and function degree of each attribute in surrounding rock classification.The forecasting method for surrounding rock classification is provided,its application is proved feasible.Separate classification forecasting of common disconstinuities and faults will bring about better results.