CN-121995491-A - Method and device for constructing seismic denoising data set

CN121995491ACN 121995491 ACN121995491 ACN 121995491ACN-121995491-A

Abstract

A method and device for constructing a seismic denoising data set comprise the steps of acquiring actual seismic data in a to-be-processed seismic work area, selecting at least two seismic channel data from the actual seismic data, determining a theoretical first arrival line of the actual seismic data, determining a noise area, calculating relative energy and absolute energy according to the actual seismic data, determining an effective signal sample area according to the relative energy and the absolute energy, and building the seismic denoising sample data set according to the noise area and the effective signal sample area.

Inventors

WANG WEI
GAO JIANHU
CHANG DEKUAN
CHEN DEWU
HE RUN
HE DONGYANG

Assignees

中国石油天然气集团有限公司

Dates

Publication Date: 20260508
Application Date: 20241106

Claims (12)

1. A method of constructing a seismic denoising data set, the method comprising: acquiring actual seismic data in a seismic work area to be processed; selecting at least two seismic channel data from the actual seismic data, determining a theoretical first arrival line of the actual seismic data, and determining a noise area; calculating relative energy and absolute energy according to the actual seismic data, and determining an effective signal sample area according to the relative energy and the absolute energy; And establishing a seismic denoising sample data set according to the noise region and the effective signal sample region.
2. The method of constructing a seismic denoising data set of claim 1, The selecting at least two seismic trace data from the actual seismic data to determine a theoretical first arrival line of the actual seismic data and to determine a noise area comprises: Selecting at least two seismic channel data from the actual seismic data, and acquiring point information of at least two points along a first arrival of the seismic data from the seismic channel data; calculating offset distances between two data points according to the seismic recording time and the channel sequence number of the two points; calculating self-excitation self-receiving time and formation speed according to the offset distance between two data points; and calculating theoretical first arrival time points of all channels in the actual seismic data according to the self-excitation self-receiving time and the stratum speed.
3. The method of constructing a seismic denoising data set of claim 2, The point information includes a seismic recording time and a channel sequence number.
4. The method of constructing a seismic denoising data set of claim 3, The calculating the offset distance between the two data points according to the seismic record time and the channel sequence number of the two points comprises the following steps: Determining the coordinates of the channel excitation point and the coordinates of the channel receiving point according to the channel serial number corresponding to each data point; Calculating corresponding offset distances by using an offset distance formula according to the coordinates of the channel excitation points of the two data points and the coordinates of the channel receiving points; wherein, the offset formula is: where x is the offset between the two points, (sx, sy) is the coordinates of the excitation point of the track and (rx, ry) is the coordinates of the receiving point of the track.
5. The method of constructing a seismic denoising data set of claim 2, The calculating the self-excitation self-receiving time and the formation speed according to the offset distance between two data points comprises the following steps: acquiring a relation between offset and time; Determining self-excitation self-receiving time and stratum speed according to offset distance and track sequence number data between two data points by using a relation formula of the offset distance and time; The relation formula of the offset distance and the time is as follows: Where x represents the offset corresponding to the data point, ti represents the time of the seismic recording, the trace number data are (x 1 ,ti 1 ) and (x 2 ,ti 2 ),t 0 is the self-excitation time, v is the formation velocity, x 1 is the offset of the first data point, x 2 is the offset of the second data point, ti 1 is the time of the seismic recording of the first data point, and ti 2 is the time of the seismic recording of the second data point.
6. The method of constructing a seismic denoising data set of claim 5, After the self-excitation self-recovery time and the formation speed are calculated according to the offset distance between the two data points, the method further comprises the following steps: judging whether theoretical first arrival lines of all the seismic channels are above actual first arrival time according to the theoretical first arrival time points; If the theoretical first arrival line is below the actual first arrival time, reselecting the two seismic trace data to recalculate the theoretical first arrival line; If the theoretical first-arrival line of all the seismic traces is above the actual first-arrival time of the seismic traces, determining the data before the theoretical first-arrival line as a noise area.
7. The method of constructing a seismic denoising data set of claim 1, Said calculating relative and absolute energies from said actual seismic data and determining an effective signal sample area from said relative and absolute energies, comprising: Respectively calculating the energy before the first arrival and the energy of the whole seismic channel according to the actual seismic data; calculating relative energy and absolute energy according to the energy before the first arrival and the energy of the whole seismic channel; Judging whether each seismic trace is an effective signal seismic trace or not by utilizing a preset relative energy threshold or absolute energy threshold according to the relative energy or the absolute energy; After the judging steps are executed for all the seismic channel data, if the effective signal seismic channels are continuous and the number of the effective signal seismic channels exceeds the minimum area threshold value, determining the current area as an effective signal sample area.
8. The method of constructing a seismic denoising data set of claim 7, The step of respectively calculating the energy before the first arrival and the energy of the whole seismic channel according to the actual seismic data comprises the following steps: energy e fb before first arrival: e fb ＝max{|d i |,i=0,1,...,T′}; Wherein e fb is the energy before the first arrival, T' is the time of the first arrival line of the theory; Energy e all of the entire seismic trace: e all ＝max{|d i |,i=0,1,...,T} Where e all is the energy of the entire seismic trace, T is the maximum recording time of the trace, ||represents taking the absolute value, and d i represents the value of the seismic sampling point at time i.
9. The method of constructing a seismic denoising data set of claim 7, The relative energy e r is: in the above formula, e r is the relative energy, e fb is the energy before the first arrival, and e all is the energy of the whole seismic trace; the absolute energy e a is: e a ＝e fb In the above formula, e a is absolute energy, and e fb is energy before first arrival.
10. The method of constructing a seismic denoising data set of claim 9, And determining whether each seismic trace is a valid signal seismic trace according to the relative energy or the absolute energy by utilizing a preset relative energy threshold or absolute energy threshold, including: Judging according to the relative energy or absolute energy by combining a judgment rule of a preset relative energy threshold or absolute energy threshold; if the relative energy corresponding to the seismic trace is smaller than the relative energy threshold value or the absolute energy corresponding to the seismic trace is smaller than the absolute energy threshold value, determining the seismic trace as an effective signal seismic trace; wherein, the judging rule is as follows: Where f is a signature of an active signal trace, 1 represents an active signal trace, 0 represents an inactive signal trace, v r is a relative energy threshold, and v a is an absolute energy threshold.
11. An apparatus for constructing a seismic denoising data set, comprising a memory for storing a program for constructing a seismic denoising data set, and a processor for reading and executing the program for constructing a seismic denoising data set, and executing the method as claimed in any one of claims 1 to 10.
12. A computer readable storage medium having stored thereon a data processing program for execution by a processor of the method of constructing a seismic denoising dataset according to any of claims 1-10.

Description

Method and device for constructing seismic denoising data set Technical Field The present disclosure relates to the field of geophysical prospecting for oil and gas, and more particularly, to a method and apparatus for constructing a seismic denoising data set. Background Aiming at a plurality of problems of complex surface conditions, complex geological conditions, hidden oil and gas reservoirs and the like in domestic oil and gas exploration, how to recover effective seismic data from complex original seismic data becomes a fundamental work which is important for solving the problems. The earthquake background has strong noise energy, seriously interferes with the representation capability of the earthquake data to the underground geological information, and even produces false images to lead to misleading of oil and gas exploration. The background noise of the earthquake exists at each time of earthquake collection, and is distributed before and after the first arrival wave of the earthquake arrives. Background noise sources are various, including various non-seismic source related noises such as wind blowing, animal activities and the like, and various seismic source related noises such as adjacent gun interference, secondary seismic sources and the like, so that the method has the characteristics of strong and weak energy, wide frequency band distribution, coexistence of coherence and random characteristics, complex probability characteristics (non-Gaussian, non-stable) and the like, and has great suppression difficulty. In addition, with the increasing popularity of the 'two wide one high' acquisition technology, the volume of the seismic data is exponentially increased, and the efficient processing of massive seismic data also provides new challenges for the denoising technology. Therefore, how to efficiently and accurately compress complex seismic noise and recover effective signals becomes one of key technologies to be solved in the field of oil and gas exploration. In the field of seismic noise suppression, deep learning is adopted to suppress complex seismic noise, but the deep learning is a data-driven algorithm, and the completeness and the authenticity of a training data set determine the denoising effect to a great extent. However, seismic data is a seismic signal acquired by an acquisition instrument in a complex environment, and it is difficult to construct a sample tag dataset of clean signals. Therefore, how to implement a method for constructing a seismic denoising data set on the premise of using actual data is a problem to be solved. Disclosure of Invention The application provides a method and a device for constructing a seismic denoising data set, which construct a high-precision data set containing denoising data-noise data on the premise of using actual data, and the denoising network model trained by using the data set greatly improves denoising precision compared with the denoising network model trained by using analog data. In a first aspect, the present application provides a method of constructing a seismic denoising dataset, the method comprising: acquiring actual seismic data in a seismic work area to be processed; selecting at least two seismic channel data from the actual seismic data, determining a theoretical first arrival line of the actual seismic data, and determining a noise area; calculating relative energy and absolute energy according to the actual seismic data, and determining an effective signal sample area according to the relative energy and the absolute energy; And establishing a seismic denoising sample data set according to the noise region and the effective signal sample region. Alternatively, the process may be carried out in a single-stage, The selecting at least two seismic trace data from the actual seismic data to determine a theoretical first arrival line of the actual seismic data and to determine a noise area comprises: Selecting at least two seismic channel data from the actual seismic data, and acquiring point information of at least two points along a first arrival of the seismic data from the seismic channel data; calculating offset distances between two data points according to the seismic recording time and the channel sequence number of the two points; calculating self-excitation self-receiving time and formation speed according to the offset distance between two data points; and calculating theoretical first arrival time points of all channels in the actual seismic data according to the self-excitation self-receiving time and the stratum speed. Alternatively, the process may be carried out in a single-stage, The point information includes a seismic recording time and a channel sequence number. Alternatively, the process may be carried out in a single-stage, The calculating the offset distance between the two data points according to the seismic record time and the channel sequence number of the two points comprises the following steps: Dete