复小波频谱分析方法及在储层预测中的应用

针对短时窗傅立叶变换频谱分析方法存在窗函数选取不确定、分辨率固定、频谱分析精度较低等问题,采用了复小波分析方法对地震信号进行时频分析,选择复Morlet小波作为分析小波,对地震信号进行复小波变换,并将尺度域转换到频率域,获得了不同频率的时频特征水平切片;用理论模型说明了复小波频谱分析精度明显高于傅立叶变换的频谱分析.对新疆某油田实际资料的处理结果表明,复小波频谱分析能细致地反映储层的分布、位置及形态,其效果优于短时窗傅立叶变换方法.

The Spectrum Decomposition with Complex Wavelet Transforms and Its Applications in Reservoir Prediction

Short-time Fourier transform time-frequency analysis method has several problems such as uncertain window function selection, single resolution, lower accuracy of analysis of the spectrum. The paper uses complex wavelet to analyze the seismic signal for time-frequency. Morlet complex wavelet taken as the one for analysis, different frequency slices of the time-frequency characteristics are acquired by transforming seismic signals into complex wavelets and the scale domain into frequency domain. Processing of the theory model with the complex wavelet time-frequency analysis shows a higher accuracy than that with the short-time Fourier transform time-frequency analysis. The same methods were utilized on processing of the seismic data in an oilfield in Xinjiang. The results showed that time-frequency analysis with complex wavelet transforms can exactly describe reservoir distribution, location and shape.