【摘要】 常用的谱分解方法因时窗固定,使时频分辨率不能变化;谱分解生成的若干共频率谱数据体不仅占用大量存储资源,而且使解释工作量繁重。为克服这两个问题,构造了一种小波函数可调整的广义S变换,它可根据地震信号的频率自适应地调整分析时宽,获得较理想的时频分辨率。通过研究储层的几种顶底反射系数组合在广义S变换域中的时频响应,发现其振幅极值出现的频率位置与储层厚度及反射系数的符号直接相关,利用这一规律提出了基于广义S变换的地震资料高效时频谱分解方法,从谱分解后的若干共频率数据体中抽取振幅极大值的频率数据合成一个谱极大值数据体,减少了数据的存储量,提高了信噪比。在实际资料处理中用于分析岩性、地质构造和砂岩储层的空间展布,取得了良好效果。更多还原

【Abstract】 Common-used spectral decomposition method made the time-frequency resolution cannot be variation because of fixed window; several common-spectral data volumes generated from spectral decomposition not only occupied a great deal of storage resource, but also made the interpretation become workload. In order to overcome these two problems, a generalized S transform is constructed, in which the wavelet function can be adjusted and the temporal width can be adjusted and analyzed adaptively according to the frequency of seismic signal. It is discovered by studying the time-frequency response in generalized S transform domain for several top and bottom reflectivity combinations of reservoir that the frequency position appearing amplitude extreme is directly correlative with the thickness of reservoir and the sign of reflectivity, which is used to present the method of high-efficient time-frequency spectral decomposition of seismic data based on generalized S transform. The method extracts the frequency data with maximum amplitude from several post-spectral-decomposition common frequency data volume to synthesize a data volume with maximum spectrum, reducing the data storage volume and improving S/N ratio. The method is used to analyze the lithology, geologic structure and spatial distribution of sand body reservoir in practical data processing, achieving good results.更多还原