优选多地震属性线性回归方法在预测煤层厚度中的应用

The Application of Optimized Multiple Seismic Attributes Linear Regression Method To Predicting Coal Seam Thickness

采用相关性回归分析和聚类分析优选出三种地震属性:瞬时频率算术平均值、总振幅、总能量。应用线性回归方法建立了瞬时频率算术平均值、总振幅、总能量和煤层厚度之间的回归模型。通过对三元三次回归模型与三元一次、三元二次回归模型预测结果的对比分析,选用三元三次回归模型预测了山西宁武某煤矿的煤层厚度。

Using linear regression analysis and cluster analysis to optimize three seismic at- tributes.computed instantaneous frequency, sum of amplitudes, sum of magnitudes. Linear regression method is applied to establish regression model among computed instantaneous frequency, sum of amplitudes, sum of magnitudes and the coal seam thickness. After com- paring the ternary cubic regression model with ternary quadratic regression model and ter- nary linear regression model, ternary cubic regression model is chosen to predict the coal seam thickness of one coal mine in Ningwu, Shanxi.