For accurately predicting the mine coal floor failure depth, based on the multiple mining area coal floor failure depth of the measured data, five indexes of the mining depth, dip angle of coal seam, mining thickness, face length and floor ability to resist damage were selected as the main control factors affecting the coal floor failure depth, and the principal component regression method was used to establish the prediction model. Finally, the reliability of the model was verified by reserved test samples. The results showed that the principal component regression method could effectively predict the failure depth of coal seam floor. The model not only had good fitting ability, but also had good prediction ability for new samples. The root mean square error of test samples was 1.206 0 m, the mean relative error was 10.751 7%, and the mean absolute error was 0.996 6 m. This study provided a new method for predicting the depth of mining failure of coal seam floor.
刘 伟. 基于主成分回归的煤层底板采动破坏深度预测模型研究[J]. 煤炭与化工, 2026, 49(1): 91-93，115..
Liu Wei. Study on prediction model of mining failure depth of coal floor based on principal component regression. CCI, 2026, 49(1): 91-93，115..
