基于Chirplet变换在地震勘探时频域地层旋回变化识别分析

Analysis of Strata Change in Time and Frequency Domain of Seismic Exploration Based on Chirplet Transformation

针对地震波信号通常为非平稳信号,传统的傅里叶变换无法分析非平稳信号,同时不能反映局部时频分布关系的问题,依次介绍了短时傅里叶变换、连续小波变换以及Chirplet变换在一次单分量、二次单分量、多分量瞬时频率分析非平稳信号的优劣差异,利用修改核函数对于Chirplet变换进行一定改进,有效提高了待分析信号的时频域分辨率。随着石油勘探开发难度的逐渐增大,对于薄互层的层间储层预测问题逐渐成为研究焦点,旋回型薄互层是其中的典型模式之一,同时也是层序地层的一种沉积组合。改进的Chirplet变换在实际应用中相较于原始Chirplet变换以及传统小波变换在研究沉积旋回分析更能清晰刻画地层旋回变化趋势,对后续解决一般性薄互层沉积旋回预测问题具有重要的研究意义。

In response to the problem that seismic wave signals are usually non-stationary,traditional Fourier transform cannot analyze non-stationary signals and cannot reflect the local time-frequency distribution relationship.The advantages and disadvantages of short-term Fourier transform,continuous wavelet transform,and Chirplet transform in analyzing non-stationary signals at single component,second single component,and multi-component instantaneous frequencies were introduced in sequence.The modified kernel function was used to improve the Chirplet transform,effectively improving the time-frequency resolution of the signal to be analyzed.With the increasing difficulty of oil exploration and development,the prediction of thin interbedded reservoirs has gradually become a research focus.Cyclic thin interbedded reservoirs are one of the typical models and also a sedimentary combination of sequence stratigraphy.Compared with the original Chirplet transform and traditional wavelet transform,the improved Chirplet transform can more clearly depict the trend of stratigraphic cycle changes in practical applications,and has important research significance for solving the prediction problem of general thin interbedded sedimentary cycles in the future.