基于高阶统计量的高分辨率时频分析方法

High resolution time-frequency analysis method based on high-order statistics

时频分析是地球物理勘探领域信号分析和处理的关键技术,通过时频分析可以进行地震旋回的划分以及等时沉积界面的识别,然而传统的地震信号时频分析方法存在分辨率低、分析结果不直观等缺点。在分析现有时频方法特点的基础上,提出了基于高阶统计量的高分辨率时频分析方法,该方法能够有效反映时变信号的频谱特性,具有分辨率高、分析结果客观准确等特点。系统阐述了基于高阶统计量的时频分析方法的基本原理,并采用正演模型分析验证了该方法的有效性,将其应用于东营凹陷北带砂砾岩体旋回与沉积期次的划分,取得了良好的效果。

Time-frequency analysis is a key technology of signal analyzing and processing in geophysical exploration. It can be used in seismic cycles dividing and isochronous depositional interface identifying. However,the traditional time-frequency analysis methods have many shortcomings of low resolution and indirect result. High-order statistics time-frequency analysis method was proposed based on analyzing the present methods. The new method can reflect the signal spectral characteristics effectively with high resolution and accurate results. Basic principle of the method has been explained ,and the validity of the method has been verified by forward model. This method has been applied in division of cycles and depositional stages of glutenite in the north slope of the Dongying Depression,which has acquired satisfactory effect.