利用模式识别方法解释微小断层

Minor fault interpretation using pattern recognition

本文介绍了利用模式识别方法（包括模糊综合评判、模糊模式识别、模糊聚类分析、线性模式识别及模糊分段线性模式识别等五种）解释煤层中的微小断层。该方法充分利用地震波动力学特征参数（包括振幅、频率、相位、波形曲率等），将模糊数学理论与模式识别技术应用到地震资料的构造解释中。通过两个矿区的地震资料处理和重新解释，对小断层的分布情况进行了推测，查明了一批新断点（垂直落差均在5m左右），其中部分已被井下地质资料所验证，取得了令人满意的地质效果。

We detail how to interpret minor faults in coal seams by using pattern recognition technique, including fuzzy multiparameter decision, fuzzy pattern recognition,fuzzy clustering analysis,linear pattern recognition and fuzzy segment linear pattern recognition. Characteristic parameters of seismic wave mechanics (including amplitude,frequency,phase and curvature of waveform),fuzzy theory and pattern recognition technique are applied to structure interpretation of seismic data. Consequently, many minor faults (the throw is about 5m)in two prospects were newly discovered by reinterpreting the processed seismic data,some of the minor faults have been confirmed by geological data in coal mining,thus resulting in satisfactory geological effect.