基于多频率分量的叠前AVO多尺度反演方法

Research on Pre-stack Multi-scale AVO Inversion in Frequency Domain Based on Multi-frequency Components

为提高频率域反演的分辨率和抗噪能力,减小频率域反演对初始模型的依赖程度,发展了多分量频率域叠前AVO弹性参数反演方法。考虑地震信号在频率域中不同频率成分可实现自动解耦,使得反问题可通过多分量迭代的途径求取。首先,利用贝叶斯公式将多频率分量联系起来,并结合稀疏约束先验提高反演算法抗噪能力。此外,采用超低频平滑模型提高多频率分量AVO反演算法的稳定性和可靠性。然后,模型测试综合验证了该方法的高分辨率特征和横向连续性,获得与理论模型吻合度较高的高分辨率反演结果。最后,实际资料表明该方法能够稳定的从叠前地震数据中获取与测井数据相吻合的高分辨率弹性参数信息,并可有效的运用到储层含油气性质检测中。

In order to improve the resolution and anti-noise ability of frequency-domain inversion,a robust pre-stack AVO inversion method in the frequency domain was developed based on Multi-frequency components.And the model-dependent problems of conventional inversion method can be solved effectively.At first,the object function was built by Bayesian framework since the seismic signal can be decoupled automatically.Furthermore,the low-frequency constraints of the model parameters were used to improve the stability and reliability of the inversion.And then,the synthetic tests were conducted to illustratethe effectiveness of our method.Compared with conventional time domain inversion,the proposed method had superior resolution and the lateral continuity of inversion result could be preserved well.Most of all,the estimation result could be in high agreement with the synthetic geologic model.Finally,applications on field seismic indicated that our method could obtain stable results of elastic parameters from pre-stack seismic data in frequency-domain.And the reliability of our estimation was demonstrated by real logging data.