论文部分内容阅读
从地震属性分析入手,提出了用于煤层顶板泥岩百分比含量预测的地震属性分析方法和BP人工神经网络岩性预测方法.以淮南矿区潘东西四采区三维地震勘探区为依托,优选出平均瞬时相位、主频序列1,能量半衰时和主频斜率等4种地震属性作为13-1煤层顶板岩性预测分析的基本参数,结合已知钻孔资料,建立了煤层顶板泥岩百分比含量BP人工神经网络预测模型,运用训练好的网络对研究区13-1煤层顶板泥岩百分比含量进行了预测分析.结果表明,BP神经网络模型具有极强的非线性逼近能力,能真实反映煤层顶板岩性与地震属性之间的非线性关系,预测结果与实测值之间误差小,相对误差一般小于10%,地震属性可以用于煤层顶板岩性分布预测.
Starting from the analysis of seismic attributes, the seismic attribute analysis method and BP artificial neural network lithological prediction method are proposed for predicting the percentage content of mudstone in the roof of coal seam.Based on the three-dimensional seismic exploration area of Pan Dong Xi No.4 mining area in Huainan Mining Area, the average instantaneous Phase, frequency sequence 1, energy half-decay time and slope of main frequency as the basic parameters of 13-1 coal seam roof lithology prediction and analysis, combined with the known drilling data, the percentage of mudstone on the roof of coal seam is established BP artificial Neural network prediction model, the prediction and analysis of the percentage of mudstone in the roof of 13-1 coal seam in the study area is carried out by using a well-trained network.The results show that the BP neural network model has strong nonlinear approximation ability, which can truly reflect the lithology and The nonlinear relationship between seismic attributes, the error between the predicted result and the measured value is small, the relative error is generally less than 10%, and the seismic attributes can be used to predict the lithologic distribution of coal seam roof.