基于灰色理论-BP神经网络的煤层瓦斯含量预测研究

Research on Prediction of Coal Seam Gas Content Base on Gray Theory and BP Neural Network

煤层瓦斯含量是矿井瓦斯灾害防治的主要参数之一,影响其分布特征的地质因素有很多。利用灰色理论的灰色关联分析法对选取的8个影响煤层瓦斯含量的地质因素进行了分析,筛选出断距、埋深、基岩厚度以及挥发分4个主要影响因素,并将其作为BP神经网络模型的输入端建立了煤层瓦斯含量预测模型。对该预测模型进行训练与仿真检验,并与传统的多元线性回归预测方法进行比较分析。

The coal seam gas content is regarded as one of the main parameters to prevent the mine gas disaster,and its distribution characteristics is affected by many geological factors.For research purpose,the grey correlation analysis method to analyze 8 geological factors,such as depth,the components of water,coal thickness,etc are adopted,through this method,the displacement of fault,depth,the bedrock thickness and volatile matter as the main factors are filtered,and than,a coal seam gas content prediction model based on gray theory and BP neural network was built by making that 4 main factors as the inputs of BP neural network.The prediction model was make a compare with the multivariable regression model by leaning,training and simulation test.