【摘要】 时频分析方法是地震数据处理和解释领域的一个重要工具.本文基于最小二乘原理和正则化方法,提出了约束最小二乘时频分析方法.该方法是在有限支撑、非正交基函数空间下,通过迭代整形正则化方法求取信号傅里叶系数的一种时频分析方法.该方法是短时傅里叶变换时频分析方法的延伸和发展,具有较高的时频分析精度,能较准确地识别信号的频谱特征.理论分析和实际资料应用表明,该方法能够较精确的刻画储层的位置和边界,能够进一步提高储层预测的精度.更多还原

【Abstract】 Time-frequency analysis is an important technical tool in seismic data analysis area.In this paper,we proposed the least square constrained method(LSCM)based on the least square and shaping regularization principles.This new method intends for computing the Fourier coefficients of interest signal under finite supported and non-orthogonal base space according to the least square and regularization principles.This is an extension of the short time Fourier transform(STFT).Comparisons with STFT using synthetic signals and real seismic data validates that the proposed method has better timefrequency resolution,can better identify the reservoir’s boundaries and locations,further improve the accuracy of reservoir prediction.更多还原