摘要
基于经典Hilbert变换计算信号瞬时参数（如瞬时频率等），当信号中噪声较强时计算结果不能很好地刻划有效信号特征．本文提出了用小波变换求能量有限实信号对应的解析信号的一个定理，在此基础上给出了用小波变换计算信号瞬时参数的算法．理论分析及模型算例结果表明，本文提出的方法计算精度高且有较强的抗噪声能力．对地震记录的褶积模型，深入地分析了不同尺度下地震记录小波变换结果及其对应的瞬时参数含义，这对实际应用有重要意义．
IP (instantaneous parameters such as instantaneous frequency) of a signal computed with HT (Hibert transform) method usually do not well characterize the signal when itS S / N is poor. So a theorem is proposed for the calculation of the analytic signal of a real-veined signal in L2(R). Based on the theorem, a wavelet-based method (called WT method). of computing the IP is given. The numerical simulation results and theoretical analysis show that the WT method has higher precision and antinoise capacity than the HT method does. On the basis of the convolutional model in the analysis of reflection seismogram, the meaning of wavelet transform of seismogram and corresponding IP at different scales is investigared. TIs is very important to practical seismic din processing and interpretation using wavelet transform.
出处
《地球物理学报》
SCIE
EI
CAS
CSCD
北大核心
1997年第6期821-832,共12页
Chinese Journal of Geophysics
基金
国家自然科学基金
