Abstract:To solve the uncertain problems in rockburst prediction, uncertain average graded analysis method is applied to predict the level of rockburst. Uncertain average graded analysis model is first established for the prediction of possibility and the level of rockburst. Four indexes such as the maximum tangential stress of the cavern wall, uniaxial compressive strength, uniaxial tensile strength, and the elastic energy index are selected as the factors for synthetic evaluation of rockburst. The indexes criteria are determined by averages of each graded sample. The intensity of rockburst is divided into four grades that are considered as four normal populations in uncertain average graded analysis. Uncertain functions are obtained by indexes criteria. Uncertain average graded analysis model were obtained through training a series of underground rock projects both at home and abroad. The faith judgment criterion is used to judge which grade the evaluating sample belongs to. The processes of the proposed method are given. The proposed model is verified by the back substitution method. Results show that the back-test-correct rate is high. The proposed model is applied to two practical projects of tunnel and mine. Cases study shows that calculation results are consistent with actual conditions. The proposed model could be used widely in practical engineering.