Analysis of coal rib fracture depth using numerical modeling and artificial neural network

The failure of coal ribs is a major hazard in underground coal mines. Researchers from the National Institute for Occupational Safety and Health (NIOSH) have been working on the development of an engineering-based rib control method. A user-friendly standalone application, called Design of Rib Support (DORS), is being developed to ease the calculations of rib support. DORS can estimate the Primary Rib Support Density (PRSD, ton/ft2) for development loading. However, it still needs to take account of another key factor, bolt length, in rib support design. The bolt length can be estimated from the rib fracture depth, and this study focuses on the coal rib fracture depth analysis with numerical modeling and artificial neural network. A total of 331 FLAC3D simulations with different mining scenarios and rib conditions were conducted for this study. Machine learning was then used to predict coal rib fracture depth under different conditions. Based on the numerical simulations, the parameters for mining scenarios and rib conditions were collected as features for machine learning, and coal rib fracture depths were collected as the target to predict. Good data correlation (r-square) was obtained with the use of an artificial neural network. The