双向预测法压制线性干扰波和多次波

Strong linear interferences and multiples suppression based on dual prediction model.

通常干扰波压制技术是利用有效波与干扰波的频率差异和视速度差异来分离和压制干扰的。利用视速度差异的f-k滤波技术,虽然有很好的压制效果,但当有效波与干扰波的视速度范围有重叠时,二者不能完全分离;此外,由于是多道处理,在压制干扰的同时,也产生了人工噪声,降低了分辨率。双向预测法利用拉冬变换同时预测有效波和干扰波的模型,以约束压制函数的计算,然后在单道记录上压制干扰波,能时变和空变设计滤波器,有很强的信噪分离能力,不会产生人工噪声。理论记录和实际地震数据的处理结果表明,该方法是一种分离和压制干扰波能力很强的高保真压制干扰波方法。

Most suppression techniques are based on the difference in frequency content and the apparent velocity difference between primaries and interferences. The apparent velocity filtering, such as f-k filtering, is often applied to seismic data for strong linear interferences and multiples suppression.Although the process often gives excellent results, it can sometime result in signal smoothing and distortion, produce artificial, and poor suppression when interferences and primaries can not be separated completely, the reject band of f-k filter overlay with the signal region in f-k domain.A new technique has been developed for strong linear interferences and multiples suppression. Radon transform has been introduced for getting better sepa ration of interferences and primaries. Both primaries and interferences are applied to restrain suppressing function calculation.It suppresses interferences and multiples on signal trace. The procession of synthetic and field data show that the technique has no signal distortion and smoothing and improve the suppression characteristics of filter. The result is superior to that obtained from the conventional multi trace suppression approach.