基于S变换域的时变反褶积方法研究
发布时间：2019-02-14 16:30
【摘要】：常见的提高地震资料分辨率的方法有反褶积,谱白化,小波变换等,这些方法都是以Robison提出的平稳褶积模型作为理论基础。而在实际地震勘探中,地震子波在地下传播时,由于介质的非均匀性和非完全弹性等会引起子波能量的衰减,振幅变小,子波长度变长等。就是说地震子波在地下介质传到的地震记录也是非平稳的。这就使得以平稳模型作为理论基础的提高分辨率的方法处理这些非平稳地震记录往往达不到预期的处理效果。常规反褶积方法具有很多局限性,往往要求地震子波是最小相位并且是平稳的,反射系数序列为白噪等等。S变换及其扩展(广义S变换)能够提供精细的时频谱,在地震数据处理中有着广泛的应用,如:地震记录的随机噪音衰减,面波压制,衰减参数提取,地震波初至检测,提取沉积旋回等。为了适应地震子波的时变特性,本文将常规褶积模型扩展成时变褶积模型,通过S变换对地震记录进行时频分析,估计出震源子波和衰减因子,求取时变子波,在S域进行时变反褶积,较好地适应了地震记录的时变特性。在S变换域中进行提高分辨率处理,可以补偿传播过程中衰减的地震子波能量,提高资料的分辨率,使得浅中深层的能量相对均衡。模型及实际资料的处理验证了本文方法的正确性,体现了一定的实用性。
[Abstract]:The common methods to improve the resolution of seismic data are deconvolution spectral whitening wavelet transform and so on. These methods are based on the stationary convolution model proposed by Robison. In the actual seismic exploration, when the seismic wavelet propagates underground, the attenuation of wavelet energy, the decrease of amplitude and the length of wavelet will be caused by the inhomogeneity and incompletely elasticity of the medium. That is to say, the seismic records transmitted by seismic wavelet in underground media are also non-stationary. Therefore, the method of improving the resolution based on stationary model is often unable to achieve the desired processing effect of these non-stationary seismic records. The conventional deconvolution method has many limitations. It is often required that the seismic wavelet is minimal phase and stationary, the reflection coefficient sequence is white noise, etc. The S transform and its extension (generalized S transform) can provide fine time spectrum. It has been widely used in seismic data processing, such as random noise attenuation of seismic records, surface wave suppression, attenuation parameter extraction, seismic wave first arrival detection, extraction of sedimentary cycles and so on. In order to adapt to the time-varying characteristics of seismic wavelet, the conventional convolution model is extended to time-varying convolution model in this paper. The time-frequency analysis of seismic records is carried out through S-transform, the source wavelet and attenuation factor are estimated, and the time-varying wavelet is obtained. The time-varying deconvolution in S domain is well adapted to the time-varying characteristics of seismic records. The improved resolution processing in S transform domain can compensate the attenuation of seismic wavelet energy in the propagation process, improve the resolution of the data, and make the energy of shallow, medium and deep layers relatively balanced. The validity of the method is verified by the processing of the model and the actual data, which reflects the practicability of the method.
【学位授予单位】：中国石油大学(华东)
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P631.4
本文编号：2422388
[Abstract]:The common methods to improve the resolution of seismic data are deconvolution spectral whitening wavelet transform and so on. These methods are based on the stationary convolution model proposed by Robison. In the actual seismic exploration, when the seismic wavelet propagates underground, the attenuation of wavelet energy, the decrease of amplitude and the length of wavelet will be caused by the inhomogeneity and incompletely elasticity of the medium. That is to say, the seismic records transmitted by seismic wavelet in underground media are also non-stationary. Therefore, the method of improving the resolution based on stationary model is often unable to achieve the desired processing effect of these non-stationary seismic records. The conventional deconvolution method has many limitations. It is often required that the seismic wavelet is minimal phase and stationary, the reflection coefficient sequence is white noise, etc. The S transform and its extension (generalized S transform) can provide fine time spectrum. It has been widely used in seismic data processing, such as random noise attenuation of seismic records, surface wave suppression, attenuation parameter extraction, seismic wave first arrival detection, extraction of sedimentary cycles and so on. In order to adapt to the time-varying characteristics of seismic wavelet, the conventional convolution model is extended to time-varying convolution model in this paper. The time-frequency analysis of seismic records is carried out through S-transform, the source wavelet and attenuation factor are estimated, and the time-varying wavelet is obtained. The time-varying deconvolution in S domain is well adapted to the time-varying characteristics of seismic records. The improved resolution processing in S transform domain can compensate the attenuation of seismic wavelet energy in the propagation process, improve the resolution of the data, and make the energy of shallow, medium and deep layers relatively balanced. The validity of the method is verified by the processing of the model and the actual data, which reflects the practicability of the method.
【学位授予单位】：中国石油大学(华东)
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P631.4
【参考文献】
相关期刊论文 前10条
1 曹健;赵宪生;施泽进;于静芳;;基于动态子波估计的吸收系数提取方法[J];物探化探计算技术;2011年03期
2 杨林;;地震频谱分解技术应用中有关问题的讨论[J];石油物探;2008年04期
3 刘喜武;张宁;勾永峰;李华;;地震勘探信号时频分析方法对比与应用分析[J];地球物理学进展;2008年03期
4 陈文超;高静怀;;基于改进的最佳匹配地震子波的地震资料衰减特性分析[J];地球物理学报;2007年03期
5 蔡希玲;刘学伟;王彦娟;吕英梅;聂红梅;;地表一致性统计相关分析法及其应用[J];石油物探;2006年04期
6 云美厚;丁伟;王开燕;唐付良;张晓梅;;地震资料一致性处理方法研究与初步应用[J];石油物探;2006年01期
7 马昭军;刘洋;;地震波衰减反演研究综述[J];地球物理学进展;2005年04期
8 曹孟起,周兴元,王君;统计法同态反褶积[J];石油地球物理勘探;2003年S1期
9 王君,周兴元,曹孟起;同态反褶积的改进与应用[J];石油地球物理勘探;2003年S1期
10 高静怀,陈文超,李幼铭,田芳;广义S变换与薄互层地震响应分析[J];地球物理学报;2003年04期
相关硕士学位论文 前1条
1 谭军;基于模型的变周期预测反褶积[D];中国海洋大学;2008年
，本文编号：2422388
本文链接：https://www.wllwen.com/kejilunwen/diqiudizhi/2422388.html
