【摘要】 时频分析(TF)是地震资料处理与解释中非常重要的方法之一，时频分辨率是高精度储层预测的关键参数。常规S变换及广义S变换的时频分辨率已难以满足高精度储层预测的需求。为此，将稀疏约束的思想引入TF中，在利用广义S变换参数可灵活调节的基础上，通过优化窗矩阵构建一种稀疏广义S变换方法。合成信号的对比分析结果表明，稀疏广义S变换方法能够获得时频分辨率更高、能量聚集性更好的时频分布，在高频和低频部分均能保持较高的时间分辨率。在实际地震数据的低频阴影检测中，该方法能更清楚地刻画油气储层的空间展布，有利于减小油气储层检测的多解性。更多还原

【Abstract】 Time-frequency analysis is an important method in seismic data processing and interpretation. Time-frequency resolution is the key to high-precision reservoir prediction, but the time-frequency resolution of conventional and generalized S-Transform cannot meet the needs of high-precision reservoir prediction. Therefore, this paper introduces the idea of sparse constraint into the time-frequency analysis, and constructs a sparse genera-lized S-Transform method by optimizing the window matrix based on flexibly adjusted generalized S-Transform parameters. The comparative analysis of synthesized signals shows that the sparse generalized S-Transform method can obtain the time-frequency distribution with higher time-frequency resolution and better energy aggregation, and maintain high time resolution at both high and low frequencies. In the low-frequency shadow detection of actual seismic data, the proposed method can describe the spatial distribution of oil and gas reservoirs more clearly, which is beneficial to reduce the multi-solution of oil and gas reservoir detection.更多还原