A new empirical chart for coal burst liability classification using Kriging method

Coal burst is a catastrophic event that induced by a large variety of certainty and uncertainty factors, and many methods have been proposed to evaluate the risk of this hazard. Conventional evaluation models or empirical criteria are influenced by the complex modelling process or undesirable accuracy. In this study, a total of 147 groups of coal burst records were used to establish the empirical classification model based on elastic energy index (W-et) and impact energy index (K-c). The classification boundaries of coal burst liabilities (CBLs), which was fitted to quantitatively analyze the risk level, and its distribution characteristics are displayed on 2D chart using Kriging method. Additionally, 43 groups of test samples were collected to further validate the reliability of the constructed spatial interpolation model. The results revealed that the classification performance of Kriging model outperforms other uncertainty-based method with accuracy 91%. It can be a valuable and helpful tool for designers to conduct the geological hazard prevention and initial design.

Automated summary

An empirical classification model based on elastic energy index (W-et) and impact energy index (K-c) was established to analyze the risk level of coal burst. The Kriging method was used to display the classification boundaries and distribution characteristics of coal burst liabilities (CBLs) on a 2D chart. The reliability of the spatial interpolation model was further validated using 43 test samples. Results showed that the Kriging model had a classification accuracy of 91% and outperformed other uncertainty-based methods. This model can be a valuable tool for geological hazard prevention and initial design.