基于CEEMDAN的地震信号高分辨率时频分析方法

希尔伯特黄变换(Hilbert-Huang transform,HHT)作为一种时频分析方法,在含有间断信号的信号中或存在噪声干扰的信号中进行经验模态分解(empirical mode decomposition,EMD)时会产生模态混叠效应,为得到更准确的原始信号分析结果,精确重构原始信号,以自适应噪声的总体集合经验模态分解(complete ensemble empirical mode decomposition of adaptive noise,CEEMDAN)代替EMD并与希尔伯特变换(Hilbert transform,HT)相结合,构建了一种适应于非线性非平稳信号的高分辨率时频分析方法。采用该方法对模拟信号和实际资料进行了时频分析处理,并与传统的时频分析方法的处理结果进行了对比分析,结果表明基于CEEMDAN的时频分析方法在保证重构信号准确性和稳定性的前提下,可精确地重构原始信号,并获得精确的频谱,从而具有更高的时间和频率分辨率,它既是一种地震资料精细解释的有力工具,也为后续的含油气检测提供了有力的支撑。

The application of the empirical mode decomposition (EMD),while using the Hilbert-Huang transform (HHT) method for time-frequency analysis,can easily cause modal aliasing in the presence of discontinuous signals or noise interference in seismic signals.This study adopted the complete ensemble empirical modal decomposition with adaptive noise (CEEMDAN) to replace the EMD;the CEEMDAN was hence combined with the Hilbert transform (HT),creating a new high-resolution time frequency analysis method that can be applied to nonlinear and non-stationary signals.A time-frequency analysis was conducted applying this method to synthetic signals and actual seismic data.The test results showed that the time-frequency analysis based on CEEMDAN can reconstruct the original signal and spectrum with stability and more accuracy compared with the traditional time-frequency analysis method,thus providing results at higher resolution.Overall,the CEEMDAN method could be applied as technical support to fine geological interpretations and to the monitoring of seismic data for oil and gas prospecting.