三维地震属性参数在煤层厚度预测中的应用

依托“西部煤炭资源高精度三维地震勘探技术”项目工程，对晋城某矿南翼大巷东南区5m×5m×1ms的三维地震数据体，采用三维地震属性参数预测煤层厚度及其变化规律：沿3煤层、15煤层10ms时窗提取地震属性42种，根据钻孔资料，计算出煤厚与地震属性相关系数；从中优选出相关系数大于0．35的地震属性，其中3煤层9个、15煤层10个；然后进行地震属性互相关分析，优选出与3煤、15煤层厚度相关系数较大的4种属性，建立预测煤厚的BP神经网络模型，分别选取3煤层12个、15煤层4个实测数据作为学习训练和测试样本，以钻孔地震属性作为学习样本，对网络进行训练，最终获得全区煤层厚度。经与预留钻孔成果资料对比，预测精度较高，结果可用。

Application of 3D Seismic Attribute Parameters in Coal Seam Thickness Prediction

Rely on the ＂Western China Coal Resource High Precision 3D Seismic Prospecting Technology＂ project, through the use of 3D seismic attribute parameters to predict coal seam thickness and its changing regulations of a 5m×5m×1ms 3D seismic data volume in a Jincheng coalmine south limb main roadway southeastern district： along the Nos.3 and 15 coal seams 10ms time window picked up 42 seismic attributes, based on drilling data calculated correlation coefficient between coal seam thicknesses and seismic attributes, selected seismic attributes with correlation coefficient 〉0.35, in which 9 in No.3 coal, 10 in No.15 coal; then carried out seismic attributes cross correlation analysis, selected out 4 attributes with larger correlation coefficients in the two coal seams, set up BP neural network model to predict coal seam thickness, 12 from No.3 coal, 4 from No.15 coal actual measured data selected respectively as learning, training and testing samples, borehole seismic attributes as learning samples to train the network, finally coal seam thickness in the whole area obtained. Contrasting with reserved drilling data demonstrated high predict precision, thus the results can be used.