Open AccessOPTIMIZATION-BASED RECURSIVE FILTERING FOR VIBROSEIS HARMONIC NOISE ELIMINATION
Author(s) -
М. С. Денисов,
Anton Egorov
Publication year - 2020
Publication title -
geofizičeskie tehnologii
Language(s) - English
Resource type - Journals
ISSN - 2619-1563
DOI - 10.18303/2619-1563-2019-2-23
Subject(s) - seismic vibrator , harmonics , control theory (sociology) , computer science , algorithm , acoustics , wavelet , noise (video) , mathematics , physics , control (management) , quantum mechanics , voltage , artificial intelligence , image (mathematics)
The vibroseis sources always produce harmonics, which are considered here as noise to be eliminated. An algorithm for removing such noise has been developed based on the previously obtained mathematical model of a vibroseis wavelet distorted by the harmonics. It is shown that the ideal inversion operator is recursive, i.e. is a filter with infinite impulse response. However, due to the peculiarity of the problem, this operator is successfully approximated by a filter with a short impulse response, allowing linearization of the optimization scheme. An objective being the energy of the noise attenuation result is formed. It allows application of the geometrical divergence correction despite the fact that it introduces distortions in the signal shape and the harmonics.