论文部分内容阅读
在对振幅值动态范围分布较大的地震数据进行强能量去噪处理时,针对常规方法通常会面临的阈值求取不准导致效果不理想、需要反复测试模块参数以及需要多轮迭代联合去噪才能达到预期效果等问题,本文提出了基于数据驱动的分贝准则小波域强能量振幅压制方法。与常规方法相比,该方法不直接对异常强振幅能量值进行统计分析,而是对振幅的能量级指数进行统计分析来确定去噪阈值,即分贝判定准则。本文采用小波变换在时频域选取最佳有效信号分布时窗进行阈值统计,然后分频压制,以进一步提升去噪效果。理论和实际数据测试表明,该方法能有效压制地震数据中的强能量振幅,对强能量振幅分布动态范围适应广、时窗依赖程度低、保幅性好,具有良好的应用前景。
When performing strong energy denoising on the seismic data with large amplitude dynamic range distribution, the inaccuracy of the threshold usually met by the conventional method results in unsatisfactory results, which requires repeated testing of the module parameters and the need for multiple rounds of joint de-noising In order to achieve the expected results and other issues, this paper presents a method based on data-driven decibel criterion wavelet energy amplitude suppression method. Compared with the conventional method, this method does not make a statistical analysis of the abnormally strong amplitude energy directly. Instead, the energy level index of the amplitude is statistically analyzed to determine the denoising threshold, that is, the decibel criterion. In this paper, wavelet transform is used to select the best effective signal distribution time-window in the time-frequency domain for threshold statistics and then frequency-divided and compressed to further enhance the denoising effect. Theoretical and practical data tests show that this method can effectively suppress the strong energy amplitude in seismic data, has a wide range of dynamic range of strong energy amplitude distribution, low dependence on time window, good amplitude-preserving property and good application prospect.