论文部分内容阅读
结合煤田横向预测研究的方法及特点,讨论了地震横向预测资料处理技术,目的层的正确标定,模型的正、反演,各种地震信息的提取,地震属性的综合分析方法等主要环节。采用Kohonen人工神经网络预测奥陶灰岩岩溶裂隙发育带,以及BP人工神经网络自动识别小断层,证明该煤田地震资料横向预测方法可行,可望成为煤田岩性横向预测的一种有效方法。
Combining with the methods and characteristics of coalfield lateral prediction, the paper discusses the major aspects of seismic horizontal prediction data processing technology, correct calibration of target layer, positive and negative inversion of model, extraction of various seismic information and comprehensive analysis of seismic attributes. Using Kohonen artificial neural network to predict karst fracture development zone of Ordovician limestone and BP artificial neural network to identify small faults automatically, it is proved that the lateral prediction method of seismic data in this coalfield is feasible and it is expected to be an effective method for lateral prediction of coalfield lithology.