Abstract: The complexity and diversity of surrounding rock affect the accurate classification of tunnel surrounding rocks. Therefore, the development trend of surrounding rock classification methods is presented, and various tunnel surrounding rock classification methods and research progress are summarized and reviewed. First, the conventional surrounding rock classification methods are summarized, their advantages and disadvantages are analyzed and compared, and the modification effects of conventional surrounding rock classification methods under special geological conditions and mechanized construction methods are summarized. Second, the novel theories and methods of surrounding rock classification arose in recent years are summarized and evaluated from the aspects of identification conditions and research methods. Finally, the advantages and disadvantages of various intelligent assistant decision algorithms for surrounding rock classification are compared and analyzed. The research results show the following: (1) At present, the conventional surrounding rock classification methods are still priorities in the actual project, but assistant modifications are adopted according to the different construction methods and geological characteristics to ensure the accuracy and objectivity of the surrounding rock classification. (2) The emerging surrounding rock classification theories and intelligent methods can better adapt to modern construction methods and concepts, but they are mostly used for auxiliary decisionmaking of surrounding rock classification because of their uncertain accuracy and universality.

Key words: tunnel engineering, surrounding rock classification, artificial intelligence