基于时频能量重排的改进小波变换谱分解  

Improved wavelet spectral decomposition using energy reassignment in time-frequency map

作　　者：尚帅[1] 韩立国[2] 周晨光 郭伟 孙海军[1] 宋文杰[1] 

机构地区：[1]中国石油塔里木油田分公司勘探开发研究院,新疆库尔勒841000 [2]吉林大学地球探测科学与技术学院,长春130026 [3]中国石油东方地球物理勘探有限公司研究院,涿州072750

出　　处：《地球物理学进展》2015年第5期2206-2211,共6页Progress in Geophysics

摘　　要：小波变换作为一种具有多分辨率特征的时频分析方法已被地球物理学家广泛应用,受不确定性原理影响,其时频分辨率有限.能量重排小波变换通过对小波变换能谱向其局部重心进行重新分配,进一步提高了时频谱图的精度.拟合信号和实际地震数据表明,利用该方法能够更准确地刻画信号的时频特征,有助于获得更好的解释效果.Although continuous wavelet transform（CWT）is widely used in the field of geophysics because of its multiresolution,it still produces ambiguous results due to Heisenberg uncertainty principle. Whereas, the timefrequency resolution of CWT method can be improved further when combined with reassignment technique. This paper implements a reassigned version of continuous wavelet transform,which reassigns the energy to its local gravity center.Results of synthetic and field data show this method can characterize nonstationary variations of complex signals with high resolution,making the interpretation easier.

关 键 词：小波变换 能谱重排 高分辨率 

分 类 号：P631[天文地球—地质矿产勘探]