【摘要】 在地震数据处理中,频谱分解是一个重要的信号处理工具。简要介绍了窗口Fourier变换(WFT),连续小波变换(CWT)和匹配逼近(Matching pursuit,MP)等3种频谱分析方法,比较了它们的特点,并提出一种实现匹配逼近分解的计算方法。用合成信号对这3种方法进行试算,结果表明了匹配逼近方法具有良好的时频特性,验证了本研究提出的实现匹配逼近算法的有效性。更多还原

【Abstract】 Spectral decomposition was an important tool of signal analysis in seismic data process.Windowed Fourier Transform (WFT),Continuous Wavelet Transform (CWT) and Matching Pursuit (MP) were introduced in this paper,and a new realization algorithm of matching pursuit was proposed.These three methods were tested on synthetic signals.The results show that matching pursuit has better time-frequency property than the other two.The results also verify the new realization algorithm of matching pursuit.更多还原