In the process of accurate interpretation of multi-wave seismic data,we wanted to solve the problem of multi-wave information recognition.Based on techniques of elastic wave forwarding,targeting the geological model o...In the process of accurate interpretation of multi-wave seismic data,we wanted to solve the problem of multi-wave information recognition.Based on techniques of elastic wave forwarding,targeting the geological model of a reservoir of an oil field exploration area,we used a high-order staggered-grid difference technology to simulate many shots of seismic records of nonzero offset shots,implemented multi-wave seismic data processing to acquire the CMP of P waves and converted waves,NMO traces of CCP pre stacks,including AVA information and superposition profiles.Based on the AVA calculation of the model,the layer parameters of the model and the forwarding wave field relations of the P-S wave,we also compared and studied the correspondence between P waves and converted waves.The results of our analysis show that the results from simulation and from the AVO analysis are consistent.Significant wave field differences between P waves and converted waves in the same reservoir were found,which are helpful in recognizing and interpreting the multi-wave information in this area.We made use of the multi-wave data to provide the important guidelines for reservoir prediction.展开更多
文摘基于动态时间规整的叠前道集剩余时差校正方法存在动态时间规整算法对噪声敏感,准确计算规整路径困难;算法采用逐点搬家法,直接对地震道作剩余时差校正容易引起地震波形畸变的问题。提出一种联合稀疏贝叶斯学习(Sparse Bayesian Learning,SBL)和动态时间规整(Dynamic Time Warping,DTW)的叠前道集剩余时差校正方法,采用SBL对地震道集进行稀疏表示,再利用DTW对稀疏表示结果进行剩余时差校正,处理后重构地震记录。结果表明,SBL具有良好的噪声鲁棒性,较少的局部最小值,以及全局最优解同时也是最稀疏解,稀疏分解后得到地下地层单位冲击响应,消除了子波影响,再进行时差校正就能避免波形畸变,同时实现了高保真剩余时差校正和随机噪声压制。数值模拟和实际资料处理结果表明该方法具有良好的应用效果。
基金the Doctor Research Fund for Universities of China (No.20070616004)the National High Technology Research and Development Program of China (No.2007AA060505)
文摘In the process of accurate interpretation of multi-wave seismic data,we wanted to solve the problem of multi-wave information recognition.Based on techniques of elastic wave forwarding,targeting the geological model of a reservoir of an oil field exploration area,we used a high-order staggered-grid difference technology to simulate many shots of seismic records of nonzero offset shots,implemented multi-wave seismic data processing to acquire the CMP of P waves and converted waves,NMO traces of CCP pre stacks,including AVA information and superposition profiles.Based on the AVA calculation of the model,the layer parameters of the model and the forwarding wave field relations of the P-S wave,we also compared and studied the correspondence between P waves and converted waves.The results of our analysis show that the results from simulation and from the AVO analysis are consistent.Significant wave field differences between P waves and converted waves in the same reservoir were found,which are helpful in recognizing and interpreting the multi-wave information in this area.We made use of the multi-wave data to provide the important guidelines for reservoir prediction.
