A Novel Approach to Rock Mass Quality Classification in Open-Pit Mining Using Evidence and Error-Eliminating Theory

ABSTRACT The condition of the rock mass significantly influences the design considerations in geotechnical engineering, and its classification is an important basis for selecting geotechnical engineering parameters. For the purpose of classifying rock mass quality reasonably and dependably, an evidence theory and error-eliminating theory-based model for classifying rock mass quality has been put forward. Firstly, evidence theory was employed to integrate the weights of the indices derived from various weighting approaches, and new weights were obtained that fuse information from different weighting methods, which reduced the uncertainty and fuzziness of the index weight. Secondly, according to the idea of model building based on error-eliminating theory, the level I rock mass defined in the rock mass classification was used as the standard for right and wrong discrimination, and the error values of the indexes were computed using an improved error function, which yielded the classification results for the rock mass. Finally, the established model was implemented in quality classification of ten rock masses in the East Gobi open-pit mine, Xinjiang, China. The results demonstrate that evidence theory is capable of merging weights from diverse sources, which eliminates the need to select a particular weighting method. Additionally, error-eliminating theory addresses the shortcomings associated with data normalisation. The model’s classification results are basically consistent with the classifications of the RMR, RQDandQsystem classification. The established model improves the objectivity of quality classification of rock mass and allows for broader application.