【摘要】 地震波形的总体变化是地震波振幅、频率、相位的综合反映,是重要的地震属性参数。地震波形分类技术充分利用了地震资料信息丰富的特点,采用神经网络算法把地震道形状(即波形特征)定量地刻画出来,通过对某一层地震数据逐道进行对比分类,细致地刻画地震信号的横向变化,从而得到与地质层位对应的地震相图,用于储层砂体及岩性油藏的预测。应用波形分类技术对古龙北地区葡萄花油层沉积微相进行了预测,预测结果与该区宏观沉积环境吻合,与单井微相匹配程度高,属性的细节变化符合沉积规律。利用预测结果提供了2口井位,实施钻探后均获得工业油流,新发现了较大储量规模的岩性油藏。更多还原

【Abstract】 The shape of seismic trace is the synthetic reflection of ampli- tude,frequency,and phase for seismic wave.The waveform classi- fication technology was adopted for trace-by-trace contrast classifi- cation through neural network in some geological strata on seismic section,which can portray the lateral changes of seismic signals in detail and obtain seismic phase diagram corresponding to the geo- logical stratum to predict the sand bodies and lithologic reservoirs. Waveform classification technology was used to predict the sedi- mentary microfacies of Putaohua formation in Gulongbei area.The predicted result coincides with the macro sedimentary environment of the area,the matching degree is high with sedimentary microfa- cies of a single well,and the detailed changes of properties agrees with the sedimentary rules.Two wells were deployed by predicted results and industrial oil flow was obtained after drilling.Lithologic deposit with large scale of scales was discovered recently.更多还原