A method for selecting wavelet for frequency-preserving reverse time migration
By selecting the sinc function that does not change the seismic band as the source wave in reverse time offset imaging, combined with the finite difference dispersion relationship, the imaging quality reduction caused by different factor waves in the prior art is solved, and high resolution is achieved and high fidelity reverse time offset imaging.
Patent Information
- Application Number
- CN202011595045.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2020-12-28
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2040-12-28
AI Technical Summary
The amplitude and phase changes caused by different factor waves in the prior art in counter-time offset imaging affect the imaging quality, and the cross-correlation imaging conditions will compress the imaging frequency band and reduce the resolution.
Through careful theoretical analysis, the source waves that do not change the seismic band are selected in the existing anti-time offset imaging framework, and combined with the finite difference dispersion relationship, the sinc function is used as the source waves to achieve counter-time offset imaging, ensuring that Integrity of data bands.
Without changing the data band, the fidelity and resolution of imaging are significantly improved, and the frequency band of seismic data is maintained to the greatest extent, thereby improving the quality of counter-time offset imaging.
Smart Images

Figure CN114690240B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of seismic exploration data processing, in particular to a frequency-preserving band reverse time migration wavelet selection method. Background Art
[0002] Reverse time migration is the current mainstream imaging method. In the reverse time migration imaging method, the source wavelet and the signal received by the detector need to be propagated underground respectively, and the image of the underground medium is extracted through the cross-correlation imaging condition. This inevitably requires the selection of a suitable source wavelet to reduce the amplitude and phase of the image due to different artificial modifications of the wavelet. There are also some scholars in the industry who compare the impact of different wavelets on imaging, and strive to reduce the transformation of the final imaging by human factors. So far, the main source wavelet implementation method in the industry is to extract a wavelet through data or directly give a rick or Gaussian function wavelet as the source for forward transmission. The time variation and space variation of the wavelet in the data bring great errors to the analysis of the wavelet. Moreover, even if the more accurate seismic wavelet obtained by analysis, after the cross-correlation imaging condition, the final imaging frequency band will be compressed, thereby reducing the resolution of the reverse time migration imaging. The present invention proposes a method that does not need to analyze the seismic data wavelet and maintains the frequency band of the data to the greatest extent, thereby greatly improving the fidelity and resolution of the imaging. Summary of the invention
[0003] In view of the shortcomings of the prior art, the present invention provides a frequency-preserving band reverse time migration wavelet selection method to solve the problems raised in the above background technology.
[0004] To achieve the above object, the present invention provides the following technical solutions: a frequency-band-preserving reverse time migration wavelet selection method, which, based on the existing reverse time migration imaging framework, conducts a detailed theoretical analysis of the influence of the selection of source wavelets on the imaging frequency band, provides a reverse time migration wavelet that does not change the seismic frequency band, and combines the dispersion relationship of finite differences to provide the most suitable source wavelet for reverse time migration imaging. The wavelet can be used to achieve reverse time migration imaging without changing the data frequency band, and comprises the following steps:
[0005] S1. Selection of frequency-preserving band reverse time migration wavelet;
[0006] S2, implement reverse time shift of frequency-preserving band;
[0007] S3. Test of the reverse time shift effect of the frequency-preserving band.
[0008] To further optimize the technical solution, in step S1, the selection principle of the frequency-preserving band reverse time migration wavelet is calculated based on the single-shot imaging condition of reverse time migration, and the formula used is: Among them, I(x, y, z) represents the image of the point (x, y, z) in space, u(x, y, z, t) represents the recorded wave field at the point (x, y, z) at time t, d(x, y, z, t) is the source wave field at the point (x, y, z) at time t, and T is the maximum recording time.
[0009] Further optimizing the technical solution, the imaging condition is actually a simplified form of the following two steps:
[0010]
[0011]
[0012] I(x, y, z, t) is the image of the point (x, y, z) at time t, is the wave field recording time at time t, τ is the cross-correlation time delay, formula (2) obtains the imaging result at time t through zero-delay cross-correlation of the wave field at time t, and formula (3) obtains the final single-shot imaging result by summing the images at all times.
[0013] Further optimize this technical solution, Transformed into the frequency domain: I(x, y, z, t) = ∫u(x, y, z, y, ω)d * (x, y, z, t, ω) dω, where * means taking the conjugate. That is, the conjugate of the source wave field and the receiving wave field corresponding to each moment are multiplied in the frequency domain, and then the frequency is summed. This is the classical cross-correlation imaging condition.
[0014] To further optimize this technical solution, the frequency band of the reverse time migration imaging is the result of the weighted input source wave field. Because the wave equation is a linear equation, the propagation process does not change the frequency information of the wave field. Therefore, the wave field d at time t * The requirement that (x, y, z, t, ω) is a full-pass signal is converted into the requirement that the input seismic wavelet be a δ function or a sinc function in the time domain. According to signal analysis theory, the δ function is an infinite high-frequency signal. For high-frequency signals, the numerical dispersion of the limited difference wavefield extrapolation is large, which will affect the final imaging quality. Therefore, the δ function type wavelet is not suitable as the input wavelet for reverse time migration.
[0015] To further optimize this technical solution, the spectrum of the Sinc function is the Rect function. According to the Fourier transform principle, the inverse Fourier transform of rect(2πaf) is By giving the highest frequency of the rect function in the frequency domain, the value of a is obtained, and then the corresponding sinc function form in the time domain is determined, thereby obtaining the wavelet form of the time domain reverse time migration.
[0016] To further optimize this technical solution, since the sinc function is an infinitely long signal in the time domain, a complete sinc wavelet is impossible to exist in the application. At the same time, the sinc function is a rapidly decaying signal. In actual processing, we only need to intercept a certain length of the sinc function value.
[0017] To further optimize the technical solution, in step S2, the implementation process of reverse time migration is as follows:
[0018] S21: Based on the frequency band of seismic data and the actual situation of the work area, the highest frequency of reverse time migration is set, and the seismic data is low-pass filtered to the highest frequency range;
[0019] S22: According to the set highest frequency, the a value is obtained, thereby determining the form of the sinc function and generating the input sub-wave of this operation;
[0020] S23: Perform wavefield backpropagation and forward propagation on the input data and the sinc function wavelet to perform imaging.
[0021] Compared with the prior art, the present invention provides a frequency-preserving band reverse time migration wavelet selection method, which has the following beneficial effects:
[0022] Based on the existing reverse time migration imaging framework, this frequency band-preserving reverse time migration wavelet selection method conducts a detailed theoretical analysis of the influence of source wavelet selection on the imaging frequency band, and provides a reverse time migration wavelet that does not change the seismic frequency band. In combination with the dispersion relationship of finite differences, the most suitable source wavelet for reverse time migration imaging is provided. The wavelet can be used to achieve reverse time migration imaging without changing the data frequency band, thereby greatly improving the fidelity and resolution of reverse time migration, and maintaining the data frequency band to the greatest extent, thereby greatly improving the fidelity and resolution of imaging. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] Figure 1 A schematic flow chart of a frequency-preserving band reverse time migration wavelet selection method proposed by the present invention;
[0024] Figure 2 A schematic diagram of a reverse time migration imaging profile using conventional rick wavelet for actual data of a frequency-preserving reverse time migration wavelet selection method proposed in the present invention;
[0025] Figure 3 A schematic diagram of the wavenumber spectrum corresponding to the reverse time migration imaging using conventional rick wavelet for actual data of a frequency-preserving band reverse time migration wavelet selection method proposed in the present invention;
[0026] Figure 4 A schematic diagram of a reverse time migration imaging profile using a sinc wavelet for actual data in a frequency-preserving reverse time migration wavelet selection method proposed in the present invention;
[0027] Figure 5 A schematic diagram of the wavenumber spectrum corresponding to the reverse time migration imaging using the sinc wavelet for actual data in a frequency-preserving band reverse time migration wavelet selection method proposed in the present invention. DETAILED DESCRIPTION
[0028] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0029] Embodiment 1:
[0030] Please refer to Figure 1 As shown, a method for selecting a frequency-band-preserving reverse time migration wavelet is provided. Based on the existing reverse time migration imaging framework, a detailed theoretical analysis is conducted on the influence of the selection of source wavelets on the imaging frequency band, and a reverse time migration wavelet that does not change the seismic frequency band is provided. In addition, the most suitable source wavelet for reverse time migration imaging is provided in combination with the dispersion relation of finite differences. The reverse time migration imaging can be realized by using the wavelet without changing the data frequency band. The method comprises the following steps:
[0031] S1. Principles for selecting frequency-preserving band-reverse time migration wavelets.
[0032] Starting from the single-shot imaging condition of reverse time migration:
[0033]
[0034] Among them, I(x,y,z) represents the image of the point (x,y,z) in space, u(x,y,z,t) represents the recorded wave field at the point (x,y,z) at time t, d(x,y,z,t) is the source wave field at the point (x,y,z) at time t, and T is the maximum recording time. The imaging condition is actually a simplified form of the following two steps:
[0035]
[0036]
[0037] I(x,y,z,t) is the image of the point (x,y,z) at time t, is the wave field recording time at time t, τ is the cross-correlation time delay, formula (2) obtains the imaging result at time t by zero-delay cross-correlation of the wave field at time t, and formula (3) obtains the final single-shot imaging result by summing the images at all times. Formula (2) Converted to frequency domain
[0038] I(x,y,z,t)=∫u(x,y,z,t,ω)d * (x,y,z,t,ω)dω, (4)
[0039] Among them, * means taking the conjugate. That is, the conjugate of the source wave field and the received wave field corresponding to each moment are multiplied in the frequency domain, and then the frequency is summed. This is the classical cross-correlation imaging condition.
[0040] Assume that the reflection coefficient of a certain point (x, y, z) underground is r(x, y, z), and the incident wavelet of the source collected at this point is W u (ω), ignoring the AVO effect of the wave field, we have
[0041] u(x,y,z,t,ω)=r(x,y,z)W u (ω). (5)
[0042] Similarly, the incident wave field at this point during the reverse time migration calculation is W d Substituting (ω) into the frequency domain imaging formula (4), we have
[0043]
[0044] It can be seen that the frequency band of the reverse time migration imaging is the result of the weighting of the input source wave field. To obtain an image without changing the waveform of the received wave field, it is only necessary to require W d (ω) The amplitude spectrum is 1 within the imaging frequency band, which requires d * (x, y, z, t, ω) is a full-pass filter within the effective frequency band of seismic records.
[0045] Because the wave equation is a linear equation, the propagation process does not change the frequency information of the wave field. Therefore, the wave field d at time t * The requirement that (x, y, z, t, ω) is a full-pass signal is converted into the requirement that the input seismic wavelet be a delta function or a sinc function in the time domain. According to signal analysis theory, the delta function is an infinite high-frequency signal. For high-frequency signals, the numerical dispersion of the limited difference wavefield extrapolation is large, which will affect the final imaging quality. Therefore, the delta function type wavelet is not suitable as the input wavelet for reverse time migration.
[0046] The spectrum of the Sinc function is the Rect function. According to the Fourier transform principle,
[0047] rect(2πaf) (7)
[0048] The inverse Fourier transform of
[0049]
[0050] By giving the highest frequency of the frequency domain rect function of formula (7), the value of a is obtained, and then the corresponding time domain sinc function form is determined, thereby obtaining the wavelet form of time domain reverse time migration.
[0051] Since the sinc function is an infinite length signal in the time domain, a complete sinc wavelet is impossible to exist in the application. At the same time, the sinc function is a rapidly decaying signal. In actual processing, we only need to intercept a certain length of the sinc function value to achieve a better application effect.
[0052] S2, frequency-preserving band reverse time migration implementation process:
[0053] According to the above theoretical analysis, the implementation process of reverse time migration is as follows:
[0054] S21: According to the frequency band of the seismic data and the actual situation of the work area, the highest frequency of the reverse time migration is set, and the seismic data is low-pass filtered to the highest frequency range.
[0055] S22: According to the set maximum frequency, the value of a in formula (7) is obtained, thereby determining the form of the sinc function in formula (8) and generating the input sub-wave for this operation.
[0056] S23: Perform wavefield back propagation and forward propagation on the input data and the sinc function wavelet respectively, and perform imaging according to formula (1).
[0057] S3, frequency-preserving band reverse time shift effect.
[0058] Next, we use actual data from a certain place to verify the correctness of the method. Figure 2 This is an imaging section obtained by reverse time migration imaging using rick wavelet as the source for some actual data. From the imaging section, the wave number range of the image is very narrow and the longitudinal resolution of the imaging is very low. Figure 3 For the corresponding Figure 2 From the wavenumber spectrum, we can see that both low and high wavenumbers decay quickly; Figure 4 This is the imaging section obtained by reverse time migration imaging using the sinc function as the source for the actual data. It can be seen that the wave number range of the reverse time migration imaging section of the sinc function is greatly broadened, and the longitudinal resolution of the imaging is greatly improved; Figure 5 For the corresponding Figure 4 The wave number spectrum of Figure 3 The low and high wavenumbers of the sinc function wavelet spectrum are well preserved, and the imaging resolution is significantly improved.
[0059] The beneficial effects of the present invention are as follows: the frequency-band-preserving reverse time migration wavelet selection method, based on the existing reverse time migration imaging framework, conducts a detailed theoretical analysis of the influence of the selection of source wavelets on the imaging frequency band, provides a reverse time migration wavelet that does not change the seismic frequency band, and combines the dispersion relationship of finite differences to provide the most suitable source wavelet for reverse time migration imaging. The wavelet can be used to achieve reverse time migration imaging without changing the data frequency band, thereby greatly improving the fidelity and resolution of reverse time migration, and maintaining the frequency band of the data to the greatest extent, thereby greatly improving the fidelity and resolution of imaging.
[0060] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.
Claims
1. A frequency-preserving band reverse time migration wavelet selection method, characterized in that: The following steps are involved: S1. Selection of frequency-preserving band reverse time migration wavelet; S2, implement reverse time shift of frequency-preserving band; S3, frequency-preserving band reverse time shift effect test; In step S1, the selection principle of the frequency-preserving band reverse time migration wavelet is calculated based on the single-shot imaging condition of reverse time migration, and the formula used is: where I(x, y, z) represents the image of the point (x, y, z) in space, u(x, y, z, t) represents the recorded wave field at the point (x, y, z) at time t, d(x, y, z, t) is the source wave field at the point (x, y, z) at time t, and T is the maximum recording time; The input seismic wavelet is required to be a δ function or a sinc function in the time domain; the spectrum of the Sinc function is the Rect function. According to the Fourier transform principle, the inverse Fourier transform of rect(2πaf) is to obtain the value of a by giving the highest frequency of the rect function in the frequency domain, and then determine the corresponding sinc function form in the time domain, so as to obtain the wavelet form of the reverse time migration in the time domain; In step S2, the implementation process of reverse time migration is as follows: S21: Based on the frequency band of seismic data and the actual situation of the work area, the highest frequency of reverse time migration is set, and the seismic data is low-pass filtered to the highest frequency range; S22: According to the set highest frequency, the a value is obtained, thereby determining the form of the sinc function and generating the input sub-wave of this operation; S23: Perform wavefield back propagation and forward propagation on the input data and the sinc function wavelet respectively to perform imaging.
2. A frequency-preserving band reverse time migration wavelet selection method according to claim 1, characterized in that: The imaging condition is actually a simplified form of the following two steps: I(x, y, z, t) is the image of point (x, y, z) at time t, is the wave field recording time at time t, τ is the cross-correlation time delay, and formula (2) obtains the imaging result at time t through zero-delay cross-correlation of the wave field at time t. The final single-shot imaging result is obtained by summing the images at all times.
3. A frequency-preserving band reverse time migration wavelet selection method according to claim 2, characterized in that: It is converted into the frequency domain as I(x, y, z, t) = ∫u(x, y, z, t, ω)d*(x, y, z, t, ω)dω, where * represents the conjugate, that is, the conjugate of the source wavefield and the received wavefield corresponding to each moment are multiplied in the frequency domain, and then the frequencies are summed. This is the classical cross-correlation imaging condition.
