深部煤层底板破坏深度多元非线性预测模型

Multi-factor Nonlinear Prediction Model of Floor Damage Depth in Deep

基于华北型深部煤层底板破坏深度实测数据,考虑了多因素耦合的特点,运用多元非线性回归的方法构建了深部煤层底板破坏深度预测模型,与《建筑物、水体、铁路及主要井巷煤柱留设与压煤开采规程》给出的线性回归公式进行对比分析,最后结合现场实测数据进行验证。研究表明,多元非线性回归公式的误差为10.26%,远小于线性回归公式的25.99%。多元非线性预测模型给出112145工作面的煤层底板破坏深度为21.37m,与实际情况吻合,而线性回归公式为22.73m,误差相对较大。预测模型对深部煤层安全开采具有一定的参考价值。

Based on measured data of floor damage depth of North China type coal seam in deep, and the characters of multi-factorcoupling were considered, prediction model of floor damage depth in deep was built by multi-factor nonlinear regression method, com-pared with linear regression formula that put forward by ‘ Building, Water, Railway and Main Roadway Coal Pillar Layout and CoalMining Regulation’, and then verified by field data. The results showed that the error of multi-factor nonlinear regression formula wasabout 10. 26%, and more less than the results 25. 99% of linear regression formula. The coal seam floor damage depth 21. 37m of112145 working face was provided by the multi-factor nonlinear prediction model, it coincided with practical situation, but the error oflinear regression formula was more larger, the prediction model could referenced for coal seam safety mining in deep.