以卡尔曼滤波法为基础的顺序自适应反褶积算法

• 作 者: M·阿夫特波·阿拉姆
• 出 处: 石油地球物理勘探
• 发表日期: 1976年S1期
• 关键词: 算法 波形变化 反褶积 地震信号 地震脉冲 参数误差 多次反射 地层学 自然频率 卡尔曼滤波法

分析地震脉冲穿过地球剖面时的波形变化,可提供该剖面良好的地层学以及成分的资料。影响该变化的部分因素为短周期的层间多次反射和自然频率以及介质的衰减系数。本文中,这些因素被组合成有理多项式滤波器模型,并提出一种順序自适应算法,根据短时间的地震信号估算多项式的系数。讨论了算法相对于起始参数误差、假定滤波阶的误差,以及信噪比的性能。用一理论合成的例子说明了结果。

Analysis of the waveform changes as the seismic pulse traverses the Earth’s cross-section provides good stratigraphy and compositional data for that section. Part of the factors that affect this change are the short-period inter-layer multiple reflections and natural frequencies and the medium attenuation coefficient. In this paper, these factors are combined into a rational polynomial filter model and a sequential adaptive algorithm is proposed to estimate the polynomial coefficients based on short-term seismic signals. The algorithm’s error with respect to the initial parameters, the error of the filter order, and the performance of the signal-to-noise ratio are discussed. An example of a theoretical synthesis illustrates the result.