CN-115639607-B - Noise suppression method, device and medium

CN115639607BCN 115639607 BCN115639607 BCN 115639607BCN-115639607-B

Abstract

The application discloses a method, a device and a medium for suppressing noise, and relates to the field of digital signal processing. The method comprises the steps of obtaining original observation data obtained by observing seismic signals through a detector, introducing a damping operator to perform rank reduction operation on the original observation data and obtaining a matrix after rank reduction, obtaining low-rank filtering data according to the matrix after rank reduction, and filtering signals with preset frequency in the low-rank filtering data by adopting a filter so as to obtain first denoising data. In the method, a damping operator is introduced in the process of performing the rank reduction operation on the original data, and because the damping operator can weaken the artificial interference introduced by the low-rank projection of strong disturbance noise and random noise, the suppression of incoherent noise in the seismic data is realized. In addition, the application also provides a noise suppression device and a computer readable storage medium, and the noise suppression device and the computer readable storage medium have the same or corresponding technical characteristics as the noise suppression method, and have the same effects.

Inventors

HUANG WEILIN
WANG FALIANG
GAO FEI
LI JIELI
LIU WEIJIE
LI JIDONG
LIU KE

Assignees

中国石油大学(北京)

Dates

Publication Date: 20260512
Application Date: 20220909

Claims (6)

1. A method of noise suppression, comprising: Acquiring original observation data obtained by observing the seismic signals by a detector; Introducing a damping operator to perform rank reduction operation on the original observation data and acquiring a matrix subjected to rank reduction; acquiring low-rank filtering data according to the reduced-rank matrix; filtering a signal with preset frequency in the low-rank filtering data by adopting a filter so as to acquire first denoising data; after the filtering the signal with the preset frequency in the low-rank filtering data by using a filter so as to obtain first denoising data, the method further comprises: Acquiring a weighting matrix according to the low-rank filtering data and the first denoising data; Filtering the low-rank filtering data and the signals with the preset frequency in the weighting matrix by adopting the filter so as to acquire second denoising data; The damping operator is determined according to a damping factor, and the obtaining of the damping factor comprises the following steps: Acquiring an upper limit value, a lower limit value, a maximum iteration number and a current iteration number of a preset damping factor; Acquiring the damping factor according to the upper limit value, the lower limit value, the maximum iteration number and the current iteration number; after the filtering of the low-rank filtered data and the signals of the preset frequencies in the weighting matrix by the filter so as to obtain second denoising data, the method further comprises: Judging whether the current iteration times are larger than the maximum iteration times or not; if yes, acquiring the second denoising data; if not, returning to the step of introducing a damping operator to perform the rank reduction operation on the original observed data and acquiring a matrix after rank reduction; In the iterative process, the method further comprises: judging whether the current iteration times are first iteration or not; If yes, iteratively approximating the original observed data by using a quadratic fitting function; if not, iteratively approximating the original observed data using a non-quadratic fit function.
2. The method of noise suppression according to claim 1, wherein the introducing a damping operator to perform a rank reduction operation on the raw observed data and obtain a rank reduced matrix comprises: performing frequency domain transformation on the original observed data and acquiring transformed original observed data; rearranging the transformed original observation data and obtaining a matrix after Hakk conversion; And performing rank reduction operation on the matrix subjected to Hank chemical conversion according to the damping factor and the cut-off rank, and acquiring the matrix subjected to rank reduction.
3. The method of noise suppression according to claim 2, wherein the obtaining low rank filtering data from the reduced rank matrix comprises: and performing anti-diagonal averaging on the matrix subjected to the rank reduction so as to acquire the low-rank filtering data.
4. A noise suppressing apparatus, comprising: The first acquisition module is used for acquiring original observation data obtained by observing the seismic signals by the detectors; The rank reduction and acquisition module is used for introducing a damping operator to perform rank reduction operation on the original observation data and acquiring a matrix subjected to rank reduction; The second acquisition module is used for acquiring low-rank filtering data according to the matrix after the rank reduction; The filtering module is used for filtering signals with preset frequency in the low-rank filtering data by adopting a filter so as to acquire first denoising data; After the filtering is performed on the signal with the preset frequency in the low-rank filtering data by adopting a filter so as to acquire first denoising data, the method further comprises: Acquiring a weighting matrix according to the low-rank filtering data and the first denoising data; Filtering the low-rank filtering data and the signals with the preset frequency in the weighting matrix by adopting the filter so as to acquire second denoising data; The damping operator is determined according to a damping factor, and the obtaining of the damping factor comprises the following steps: Acquiring an upper limit value, a lower limit value, a maximum iteration number and a current iteration number of a preset damping factor; Acquiring the damping factor according to the upper limit value, the lower limit value, the maximum iteration number and the current iteration number; after the filtering is performed on the low-rank filtering data and the signals with the preset frequencies in the weighting matrix by adopting the filter so as to obtain second denoising data, the method further comprises: Judging whether the current iteration times are larger than the maximum iteration times or not; if yes, acquiring the second denoising data; if not, returning to the step of introducing a damping operator to perform the rank reduction operation on the original observed data and acquiring a matrix after rank reduction; In the iterative process, the method further comprises the following steps: judging whether the current iteration times are first iteration or not; If yes, iteratively approximating the original observed data by using a quadratic fitting function; if not, iteratively approximating the original observed data using a non-quadratic fit function.
5. A noise suppressing apparatus, comprising: A memory for storing a computer program; A processor for implementing the steps of the method of noise suppression according to any one of claims 1 to 3 when executing said computer program.
6. A computer readable storage medium, characterized in that the computer readable storage medium has stored thereon a computer program which, when executed by a processor, implements the steps of the method of noise suppression according to any of claims 1 to 3.

Description

Noise suppression method, device and medium Technical Field The present application relates to the field of digital signal processing, and in particular, to a method, an apparatus, and a medium for noise suppression. Background In the process of recording the seismic data, random noise (the random noise belongs to incoherent interference waves) possibly doped in the seismic data, and noise with larger intensity can form larger interference on the recording of the seismic data, so that the recorded seismic data is inaccurate, and therefore, the incoherent noise in the seismic data needs to be suppressed. At present, singular spectrum analysis (Singular Spectrum Analysis, SSA) is commonly used to suppress seismic incoherent noise. The SSA method uses the low rank characteristics of the seismic signals to treat noise suppression as a low rank reconstruction problem. However, in the process of adopting the SSA method, since low-rank projection in SSA is forced to introduce significant artificial interference to fit high-amplitude unstable interference in the least square sense, incoherent noise in the seismic data cannot be effectively suppressed by adopting the SSA method. It can be seen that how to suppress incoherent noise in seismic data is a technical problem that needs to be solved by those skilled in the art. Disclosure of Invention The application aims to provide a method, a device and a medium for suppressing noise, which are used for suppressing incoherent noise in seismic data. In order to solve the above technical problems, the present application provides a method for suppressing noise, including: Acquiring original observation data obtained by observing the seismic signals by a detector; Introducing a damping operator to perform rank reduction operation on the original observation data and acquiring a matrix subjected to rank reduction; acquiring low-rank filtering data according to the reduced-rank matrix; And filtering a signal with preset frequency in the low-rank filtering data by adopting a filter so as to acquire first denoising data. Preferably, after the filtering the signal of the preset frequency in the low-rank filtered data with the filter so as to obtain the first denoising data, the method further includes: Acquiring a weighting matrix according to the low-rank filtering data and the first denoising data; and filtering the low-rank filtering data and the signals with the preset frequency in the weighting matrix by adopting the filter so as to acquire second denoising data. Preferably, the damping operator determines according to a damping factor, and acquiring the damping factor includes: Acquiring an upper limit value, a lower limit value, a maximum iteration number and a current iteration number of a preset damping factor; And acquiring the damping factor according to the upper limit value, the lower limit value, the maximum iteration number and the current iteration number. Preferably, the introducing the damping operator to perform a rank reduction operation on the original observed data and obtain a matrix after rank reduction includes: performing frequency domain transformation on the original observed data and acquiring transformed original observed data; rearranging the transformed original observation data and obtaining a matrix after Hakk conversion; And performing rank reduction operation on the matrix subjected to Hank chemical conversion according to the damping factor and the cut-off rank, and acquiring the matrix subjected to rank reduction. Preferably, the obtaining low rank filtering data according to the reduced rank matrix includes: and performing anti-diagonal averaging on the matrix subjected to the rank reduction so as to acquire the low-rank filtering data. Preferably, after the filtering the low rank filtered data with the filter and the signal of the preset frequency in the weighting matrix so as to obtain second denoising data, the method further includes: Judging whether the current iteration times are larger than the maximum iteration times or not; if yes, acquiring the second denoising data; And if not, returning to the step of introducing a damping operator to perform the rank reduction operation on the original observed data and acquiring a matrix after rank reduction. Preferably, in the iterative process, the method further comprises: judging whether the current iteration times are first iteration or not; If yes, iteratively approximating the original observed data by using a quadratic fitting function; if not, iteratively approximating the original observed data using a non-quadratic fit function. In order to solve the technical problem, the present application further provides a device for suppressing noise, including: The first acquisition module is used for acquiring original observation data obtained by observing the seismic signals by the detectors; The rank reduction and acquisition module is used for introducing a damping operator to perform ra