利用人工神经网络法预测煤层厚度研究

我国的兖州、淮北、平顶山、焦作等矿区,普遍存在着煤层缺失、剥失、分叉、合并等现象,煤层厚度变化对煤炭开采产生很大影响.据资料统计,如果实际煤厚比设计煤厚变薄10%～20%时,煤炭产量就会下降35%～40%.煤田高分辨率三维地震勘探的开展,为解决地质问题提供了丰富的三维数据体,特别是为BP人工神经网络法研究煤层厚度提供了多种地震波属性.

Coal Seam Thickness Predicting by the Use of Artificial Neural Network Simulation

Phenomena of coal seam hiatus, denudating, splitting and merging are very often in Yanzhou, Huaibei, Pingdingshan and Jiaozuo mining areas. The variation of coal seam thickness will greatly impact the coal mining. Based on data statistics, if actual coal seam thickness is 10-20% thinner than designed, coal production will be 35-40% dropped. Coalfield high resolution 3D seismic prospecting provides with abundant 3D data volume for geologic problem solving, especially multiple seismic wave attributes for coal seam thickness study by the use of artificial neural network simulation.