Prediction of Top-Coal Caving and Drawing Characteristics Using Artificial Neural Networks in Extremely Thick Coal Seam

Based on the feature of large thickness and poor drawing characteristics in extremely thick coal seam top-coal caving method, combined with numerous practical examples analyses, the primarily six factors influence the drawing characteristics were found out which are mining depth, coal seam strength, joint crack development, parting thickness in top-coal, caving ratios, immediate roof filling coefficient. According to 45 typical top-coal caving in extremely thick coal seam samples, the prediction of top-coal caving and drawing characteristics based on artificial neural networks was established and training samples and testing samples was determined. Use SPSS statistical software training the network model. Then select No. 9 coal seam first mining area of Tiaohu mine as the application case. The drawing property was forecast according to the established network model. Application results show that the use of artificial neural networks for top-coal caving and drawing characteristic prediction is effective and feasible.