Rockburst prediction based on optimization of unascertained measure theory with normal cloud

Abstract

Abstract Rockburst is one of the common geological disasters in deep underground areas with high stress. Rockburst prediction is an important measure to know in advance the risk of rockburst hazards to take a scientific approach to the response. In view of the fuzziness and uncertainty between quantitative indexes and qualitative grade assessments in prediction, this study proposes the use of a normal cloud model to optimize the theory of unascertained measures (NC-UM). The uniaxial compressive strength (σ c), stress coefficient (σ θ /σ c), elastic deformation energy index (Wet), and brittleness index of rock (σ c/σ t) are selected as the index of prediction. After data screening, 249 groups of rockburst case data are selected as the original data set. To reduce the influence of subjective and objective factors of index weight on the prediction results, the game theory is used to synthesize the three weighting methods of Criteria Importance Through Intercriteria Correlation (CRITIC), Entropy Weight (EW), and Analytic Hierarchy Process (AHP) to obtain the comprehensive weight of the index. After validating the model with example data, the results showed that the model was 93.3% accurate with no more than one level of prediction deviation. Compared with the traditional unascertained measure (UM) rockburst prediction model, the accuracy is 15–20% higher than that of the traditional model. It shows that the model is valid and applicable in predicting the rockburst propensity level.

Keywords

• Rockburst prediction
• Normal cloud model
• Unascertained measure theory
• Confidence recognition
• Comprehensive weighting of game