论文部分内容阅读
导水裂隙带高度的预测对煤矿安全开采有重要意义,而传统回归方法未考虑因素间相关系数对预测结果的影响。选取采深、煤层倾角、煤层厚度、煤层硬度、岩层结构、顶板岩石单轴抗压强度、开采厚度和采空区斜长作为预测导水裂隙带高度的影响因素,建立基于PCA-BP神经网络的导水裂隙带高度预测模型。测试结果表明,煤层厚度对导水裂隙带高度的影响最大,其余各因素对导水裂隙带高度的影响较大,采深和开采厚度对导水裂隙带高度的影响较小;PCA-BP神经网络模型的训练速度和预测效果均优于BP神经网络模型,且最大预测误差仅为5.58%。
The prediction of the height of water-conducted fractured zone is of great significance for the safe mining of coal mines, while the traditional regression method does not consider the influence of the correlation coefficient between factors on the prediction results. Selecting the factors of deep mining, coal seam dip angle, coal seam thickness, coal seam hardness, rock strata structure, roof uniaxial compressive strength, mining thickness and goaf slant length as predictors of the height of watercured fracture zone, a PCABP neural network The prediction model of the height of water-conducting fractured zone. The test results show that the thickness of coal seam has the greatest influence on the height of water-conducted fracture zone, and the other factors have a great influence on the height of water-conducted fracture zone. The influence of mining depth and mining thickness has little effect on the height of water- The training speed and forecasting effect of the network model are better than the BP neural network model, and the maximum prediction error is only 5.58%.