A method and device for high frequency extension of seismic signals

Abstract

The application provides a seismic signal high frequency extension method and device, and belongs to the technical field of seismic data processing. Fidelity and amplitude-preserved post-stack seismic data needing high frequency extension is acquired, and a seismic signal is extracted from the acquired fidelity and amplitude-preserved post-stack seismic data; the seismic signal is transformed to a frequency domain; high frequency extension is performed on the seismic signal in the frequency domain; a wavelet is eliminated in the frequency domain, and the high frequency extended seismic signal is transformed to a time domain; well data and known geological information are used to determine the reliability of the high frequency extended data, and the seismic data is used for structure interpretation and reservoir prediction. The high frequency extension effect of the application is better than that of a conventional deconvolution seismic high frequency extension method, the high frequency component of the seismic signal is effectively widened under the premise of not reducing the signal-to-noise ratio and the lateral continuity of the seismic data, the fidelity and amplitude-preserved characteristics of the high frequency extended seismic profile are well maintained, and reliable high frequency data is provided for the structure interpretation and reservoir prediction of continental facies thin sand bodies.

Description

Technical Field

[0001] This invention relates to the field of seismic data processing technology, and in particular to a method and apparatus for high-frequency extension of seismic signals. Background Technology

[0002] As oil and gas exploration targets gradually shift from structural oil and gas reservoirs to complex oil and gas reservoirs such as lithological and unconventional ones, the identification and characterization of small-scale geological targets such as thin reservoirs and small faults have placed higher demands on the accuracy of seismic exploration. The resolution of currently acquired seismic signals is difficult to meet the actual needs of effectively identifying thin reservoirs and small faults. Improving the resolution of seismic signals is of great significance for improving the success rate of exploration of complex oil and gas reservoirs.

[0003] The reflection coefficient in seismic signals is full-bandwidth, meaning that even thin layers can be reflected as long as there is an impedance difference. However, seismic wavelets are band-limited, resulting in band-limited seismic signals with narrow bandwidth and low resolution, making it difficult to identify thin layers or thin interlayers. Improving resolution has always been a crucial aspect of seismic data processing and a significant factor limiting the accuracy of oil and gas exploration, placing increasingly higher demands on high-resolution processing technologies.

[0004] The common method for improving the resolution of seismic signals is deconvolution, which primarily works by compressing the wavelet. Deconvolution-based methods can improve the frequency of seismic signals to a certain extent. However, existing deconvolution-based high-frequency seismic spread methods struggle to effectively compress the wavelet under complex conditions, resulting in limited frequency enhancement capabilities. Furthermore, they significantly reduce the signal-to-noise ratio of seismic data, lead to poor phase axis continuity, introduce numerous false faults, and fail to maintain the fidelity and amplitude characteristics of the seismic signal, resulting in low precision in seismic structural interpretation and large reservoir prediction errors. To fully extract broadband information from seismic acquisition data while preserving its effective frequency components, more advanced and effective high-frequency seismic spread methods and technologies are urgently needed.

[0005] Chinese patent application CN113419275A discloses a high-resolution seismic processing method based on sparse dictionary learning. The method includes: upsampling and smoothing interpolation of seismic wavelet sequences; upsampling and zero-padding continuation of a formation thickness model; convolution of the two to obtain a seismic response model; obtaining a learning dictionary through channel-by-channel cosine transform; upsampling and zero-padding continuation of actual seismic data; cosine transforming the actual seismic data to obtain dictionary-learned observation data; using the reflection coefficient as the target model in the dictionary learning model; optimizing the reflection coefficient using a dual optimization method with sparse constraints based on the 1-norm to obtain the reflection coefficient; and downsampling to obtain the final reflection coefficient. This process is repeated to improve the resolution of the entire data volume. This scheme compresses the wavelet through upsampling, limiting its high-frequency extension capability, and its convolution method is computationally complex.

[0006] The existing technology has the following shortcomings:

[0007] 1. In complex situations, it is difficult to effectively compress wavelets, and the ability to increase frequency is limited;

[0008] 2. This leads to a significant decrease in the signal-to-noise ratio of seismic data, poor continuity of phase axes, and the appearance of many false fractures;

[0009] 3. It is difficult to maintain the fidelity and amplitude characteristics of seismic signals, resulting in low precision in seismic tectonic interpretation and large errors in reservoir prediction. Summary of the Invention

[0010] To address the problems existing in the prior art, this invention provides a method and apparatus for high-frequency extension of seismic signals, comprising the following steps: acquiring high-fidelity, amplitude-preserving post-stack seismic data requiring high-frequency extension; extracting seismic signals from the acquired high-fidelity, amplitude-preserving post-stack seismic data; transforming the seismic signals to the frequency domain; performing high-frequency extension on the seismic signals in the frequency domain; eliminating wavelets in the frequency domain and transforming the high-frequency extended seismic signals to the time domain; determining the reliability of the high-frequency extended data using well data and known geological information; and using the seismic data for structural interpretation and reservoir prediction. The high-frequency extension effect of this invention is superior to traditional deconvolution seismic high-frequency extension methods, providing high-resolution seismic data for fine structural interpretation and thin reservoir prediction, meeting the current needs of fine-target exploration. Without reducing the signal-to-noise ratio and lateral continuity of the seismic data phase axis, it effectively broadens the high-frequency components of the seismic signal, and the high-fidelity, amplitude-preserving characteristics of the high-frequency extended seismic profile are well maintained, providing reliable high-frequency data for structural interpretation and reservoir prediction of continental thin sand bodies.

[0011] This invention provides a method for high-frequency spread of seismic signals, comprising the following steps:

[0012] Acquire high-fidelity, amplitude-preserving post-stack seismic data that requires high-frequency extension, and extract seismic signals from the acquired high-fidelity, amplitude-preserving post-stack seismic data;

[0013] Transform the seismic signal into the frequency domain;

[0014] High-frequency extension of seismic signals in the frequency domain;

[0015] Eliminate the wavelet in the frequency domain and transform the high-frequency spread seismic signal to the time domain;

[0016] Using well data and known geological information, the reliability of high-frequency extended data is determined, and seismic data is used for structural interpretation and reservoir prediction.

[0017] Preferably, the specific method for transforming seismic signals to the frequency domain includes:

[0018] Time-domain seismic signals are represented by the following formula:

[0019] s(t)=r(t)*w(t)

[0020] The seismic signal transformed to the frequency domain is represented by the following equation:

[0021] s'(f)=F(s(t))=r'(f)·w'(f)

[0022] in,

[0023] s(t) is the time-domain seismic signal;

[0024] s'(f) is the frequency domain seismic signal;

[0025] w(t) is a time-domain seismic wavelet;

[0026] w'(f) is a frequency-domain seismic wavelet;

[0027] r(t) is the time-domain reflection coefficient;

[0028] r'(f) is the frequency domain reflection coefficient;

[0029] F represents the Fourier transform.

[0030] The seismic signal is generated by convolving a reflection coefficient sequence with a seismic wavelet. The reflection coefficient sequence represents the difference in formation impedance parameters, and the seismic wavelet is a seismic wave formed when a sharp pulse excited by an artificial source during seismic exploration gradually stabilizes after propagating a certain distance. Therefore, this invention provides the aforementioned time-domain seismic signal.

[0031] Preferably, the specific method for high-frequency extension of seismic signals in the frequency domain is as follows:

[0032] The high-frequency spread of seismic signals in the time domain is expressed by the following formula:

[0033] s hf (t)=r(t)*w(at)

[0034] The transformation of a time-domain high-frequency spread seismic signal to the frequency domain is expressed by the following formula:

[0035]

[0036] in,

[0037] s hf (t) represents the high-frequency spread seismic signal in the time domain;

[0038] s' hf (f) is the seismic signal transformed from the time-domain high-frequency spread seismic signal to the frequency domain;

[0039] w(at) is the time-domain seismic wavelet after scaling transformation;

[0040] This is the frequency domain seismic wavelet after scaling transformation;

[0041] a is the scaling factor, where a > 1.

[0042] The key to high-frequency extension of seismic signals lies in the compression of the wavelet, or the extension of the seismic wavelet spectrum. From the perspective of signal scaling characteristics, compression in the time domain corresponds to expansion in the frequency domain. Since the convolution calculation of the reflection coefficient and wavelet in the time domain of a seismic signal is complex, a Fourier transform is used to transform the convolution calculation into a product calculation in the frequency domain. This invention innovatively extends the seismic signal at high frequencies in the frequency domain, achieving the goal of compressing the wavelet. Furthermore, by eliminating the wavelet in the frequency domain, it overcomes the low signal-to-noise ratio drawback of traditional methods that obtain high frequencies through wavelet compression.

[0043] Preferably, the wavelet is eliminated in the frequency domain, and the high-frequency spread seismic signal is transformed to the time domain. Specific methods include:

[0044] The frequency of the hour window seismic signal is broadened in the frequency domain and then inverse Fourier transformed to the time domain.

[0045] The frequency of the next hourly window is then overridden until the frequency of the entire seismic signal is overridden.

[0046] Repeat the above two steps for each signal of the entire 3D seismic data to perform frequency upscaling, thereby obtaining high-resolution seismic data.

[0047] Preferably, the hour window of the Fourier transform is determined based on the time window of the seismic data.

[0048] Preferably, broadening the frequency of the hourly window seismic signal in the frequency domain and performing an inverse Fourier transform to the time domain includes:

[0049] Removing the wavelet from the high-frequency spread domain of the seismic signal yields the following formula:

[0050]

[0051] Actual seismic wavelets are time-varying and space-varying, making it impossible to obtain the true seismic wavelets. In other words, it is impossible to directly obtain the high frequency of the seismic signal by compressing the wavelets. This invention innovatively removes the wavelets in the high-frequency extended frequency domain of the seismic signal, thus eliminating the influence of the wavelets.

[0052] If the reflection coefficient of the segment within the longitudinal hour window is uniformly distributed white noise, and the reflection within the hour window is equal before and after scaling to three decimal places, then the following formula applies:

[0053]

[0054] The high-frequency spread of the seismic signal within the nth time window is expressed by the following formula:

[0055]

[0056] in,

[0057] The frequency domain reflection coefficient after scaling;

[0058] This is the frequency domain seismic signal after scaling.

[0059] T p The hour window length for the Fourier transform;

[0060] n = 1, 2, 3... indicates that the hourly window length is T. p The time window number of the captured seismic signal;

[0061] Indicates that the hourly window length is T p The frequency domain reflection coefficient of the nth time window;

[0062] The length of the hour window after scaling is T. p The frequency domain reflection coefficient of the nth time window;

[0063] Indicates that the hourly window length is T p The seismic signal with frequency spread in the frequency domain of the nth time window;

[0064] The length of the hour window after scaling is T. p The nth time window frequency domain seismic signal;

[0065] Preferably, the frequency extension is performed on the next adjacent hourly window until the entire seismic signal is extended. Specifically, this involves adding together the frequency-spread seismic signals within all time windows to obtain a complete high-frequency spread seismic signal in the frequency domain.

[0066]

[0067] Performing an inverse Fourier transform on the above equation yields the seismic signal after high-frequency time-domain expansion:

[0068] s(t) = F -1 (s) hf (f))

[0069] in,

[0070] N represents the hourly window length as T. p The total number of seismic signals captured.

[0071] s”hf (f) represents the hourly window length T. p A complete seismic signal with high-frequency extension in the frequency domain;

[0072] s(t) is s” hf (f) Time-domain seismic signal after inverse Fourier transform;

[0073] T is the total length of the seismic signal;

[0074] F -1 This represents the inverse Fourier transform.

[0075] Preferably, determining the reliability of the high-frequency extended data using well data and known geological information specifically includes:

[0076] By using well logging acoustic waves and density curves to create synthetic records and comparing them with high-frequency extended seismic data, the fidelity and amplitude of the high-frequency extended seismic data are determined.

[0077] Preferably, the use of seismic data for structural interpretation and reservoir prediction specifically includes: performing seismic structural interpretation and reservoir prediction on high-frequency extended seismic data.

[0078] Preferably, seismic tectonic interpretation and reservoir prediction on high-frequency extended seismic data includes seismic tectonic interpretation and reservoir prediction through at least one of layer tracking, fault interpretation, and formation elastic parameter inversion.

[0079] Preferably, seismic tectonic interpretation and reservoir prediction are performed through stratigraphic tracing, specifically as follows:

[0080] By synthesizing seismic records from acoustic and density data from well logging, the characteristics and location of seismic phase axes corresponding to the strata that need to be traced and interpreted are identified. The phase axes of the strata are then traced in the high-frequency extended seismic signal profile to obtain the spatial distribution characteristics of the strata.

[0081] Preferably, the seismic tectonic interpretation and reservoir prediction through fault interpretation specifically involves:

[0082] Identify whether there is a fault on the seismic phase axis corresponding to the tracked layer. The fault is the reflection feature of the underground fault. Tracking and interpreting the fault feature is called fault interpretation. The interpretation of the layer and the fault are collectively called structural interpretation.

[0083] Preferably, the seismic tectonic interpretation and reservoir prediction using formation elastic parameters specifically involves:

[0084] The formation reflection coefficient is obtained by using sparse pulse deconvolution on the high-frequency spread seismic signal.

[0085] Then, the reflection coefficient is integrated and combined with the low-frequency components of the low-frequency model of elastic parameters established by the well logging curves to realize the inversion of formation elastic parameters using high-frequency extended seismic signals;

[0086] The values ​​of the elastic parameters corresponding to the reservoir are obtained through well logging interpretation. Based on the formation elastic parameters obtained by inversion, the characteristics and range of the reservoir are delineated using the values ​​of the elastic parameters corresponding to the reservoir.

[0087] The present invention provides a high-frequency seismic signal extension device, employing any of the above-mentioned high-frequency seismic signal extension methods, comprising: a seismic signal acquisition module, a seismic signal frequency extension module, and a high-frequency extension result verification module;

[0088] The seismic signal acquisition module performs the following operations:

[0089] Acquire high-fidelity, amplitude-preserving post-stack seismic data that requires high-frequency extension, and extract seismic signals from the acquired high-fidelity, amplitude-preserving post-stack seismic data;

[0090] The seismic signal frequency extension module performs the following operations:

[0091] Transform the seismic signal into the frequency domain;

[0092] High-frequency extension of seismic signals in the frequency domain;

[0093] Eliminate the wavelet in the frequency domain and transform the high-frequency spread seismic signal to the time domain;

[0094] The high-frequency extension result verification and utilization module performs the following operations:

[0095] Using well data and known geological information, the reliability of high-frequency extended data is determined, and seismic data is used for structural interpretation and reservoir prediction.

[0096] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0097] (1) This invention avoids the traditional high-frequency extension algorithm's calculation of wavelets by removing wavelets within a small window in the frequency domain. Traditional deconvolution-based high-frequency extension methods calculate an approximate or theoretical wavelet, which fails to obtain the true time-varying and space-varying seismic wavelets in the subsurface. Furthermore, traditional compressed wavelet methods cannot satisfy the consistency of the wavelet's transverse phase, resulting in poor continuity of the seismic phase axis, low signal-to-noise ratio, and low accuracy in structural interpretation and reservoir prediction. This invention innovatively proposes a high-frequency extension formula, obtaining a novel high-frequency extension method for seismic signals. Practical application results show that the high-frequency extension effect of this invention is superior to the traditional deconvolution seismic high-frequency extension method, providing high-resolution seismic data for detailed structural interpretation and thin reservoir prediction, thus meeting the current needs of detailed target exploration.

[0098] (2) The method of obtaining high-frequency extension of seismic signals in this invention is actually a scale transformation of the hour window frequency domain seismic signal. This method can effectively broaden the high-frequency components of the seismic signal without reducing the signal-to-noise ratio and the lateral continuity of the phase axis. The high-frequency extended seismic profile is well preserved in terms of fidelity and amplitude, providing reliable high-frequency data for the structural interpretation of continental thin sand bodies and reservoir prediction. Attached Figure Description

[0099] Figure 1 (a) and (b) are schematic diagrams of an original seismic profile and its spectrum, respectively.

[0100] Figure 2 (a) and (b) are schematic diagrams of a deconvolution-processed seismic profile and a deconvolution-processed seismic profile spectrum, respectively.

[0101] Figure 3 (a) and (b) are respectively a high-frequency extended seismic profile and a high-frequency extended seismic profile spectrum diagram of an embodiment of the present invention;

[0102] Figure 4 for Figure 3 (a) Comprehensive columnar section of Zhongjing 3;

[0103] Figure 5 (a) and (b) are schematic diagrams of an original seismic profile and its spectrum, respectively;

[0104] Figure 6 (a) and (b) are respectively a high-frequency extended seismic profile and a high-frequency extended seismic profile spectrum diagram of an embodiment of the present invention;

[0105] Figure 7 This is a preliminary seismic fault prediction plane attribute map.

[0106] Figure 8 This is a planar property map for predicting fracture after high-frequency extension, according to an embodiment of the present invention.

[0107] Figure 9 A flowchart of a high-frequency extension method for seismic signals according to an embodiment of the present invention. Detailed Implementation

[0108] The following is in conjunction with the appendix Figure 1-9 The specific embodiments of the present invention will be described in detail below.

[0109] This invention provides a method for high-frequency spread of seismic signals, comprising the following steps:

[0110] Acquire high-fidelity, amplitude-preserving post-stack seismic data that requires high-frequency extension, and extract seismic signals from the acquired high-fidelity, amplitude-preserving post-stack seismic data;

[0111] Transform the seismic signal into the frequency domain;

[0112] High-frequency extension of seismic signals is performed in the frequency domain; before extension, the high-frequency signal is weak, and after extension, the energy is enhanced, allowing for the identification of more subtle details.

[0113] Eliminate the wavelet in the frequency domain and transform the high-frequency spread seismic signal to the time domain;

[0114] Using well data and known geological information, the reliability of high-frequency extended data is determined, and seismic data is used for structural interpretation and reservoir prediction.

[0115] According to a specific embodiment of the present invention, a method for transforming seismic signals to the frequency domain includes:

[0116] Time-domain seismic signals are represented by the following formula:

[0117] s(t)=r(t)*w(t)

[0118] The seismic signal transformed to the frequency domain is represented by the following equation:

[0119] s'(f)=F(s(t))=r'(f)·w'(f)

[0120] in,

[0121] s(t) is the time-domain seismic signal;

[0122] s'(f) is the frequency domain seismic signal;

[0123] w(t) is a time-domain seismic wavelet;

[0124] w'(f) is a frequency-domain seismic wavelet;

[0125] r(t) is the time-domain reflection coefficient;

[0126] r'(f) is the frequency domain reflection coefficient;

[0127] F represents the Fourier transform.

[0128] The seismic signal is generated by convolving a reflection coefficient sequence with a seismic wavelet. The reflection coefficient sequence represents the difference in formation impedance parameters, and the seismic wavelet is a seismic wave formed when a sharp pulse excited by an artificial source during seismic exploration gradually stabilizes after propagating a certain distance. Therefore, this invention provides the aforementioned time-domain seismic signal.

[0129] According to a specific embodiment of the present invention, the method for high-frequency extension of seismic signals in the frequency domain is as follows:

[0130] The high-frequency spread of a seismic signal in the time domain is expressed by the following formula:

[0131] s hf (t)=r(t)*w(at)

[0132] The transformation of a time-domain high-frequency spread seismic signal to the frequency domain is expressed by the following formula:

[0133]

[0134] in,

[0135] s hf (t) represents the high-frequency spread seismic signal in the time domain;

[0136] s' hf (f) is the seismic signal transformed from the time-domain high-frequency spread seismic signal to the frequency domain;

[0137] w(at) is the time-domain seismic wavelet after scaling transformation;

[0138] This is the frequency domain seismic wavelet after scaling transformation;

[0139] a is the scaling factor, where a > 1.

[0140] The key to high-frequency extension of seismic signals lies in the compression of the wavelet, or the extension of the seismic wavelet spectrum. From the perspective of signal scaling characteristics, compression in the time domain corresponds to expansion in the frequency domain. Since the convolution calculation of the reflection coefficient and wavelet in the time domain of a seismic signal is complex, a Fourier transform is used to transform the convolution calculation into a product calculation in the frequency domain. This invention innovatively extends the seismic signal at high frequencies in the frequency domain, achieving the goal of compressing the wavelet. Furthermore, by eliminating the wavelet in the frequency domain, it overcomes the low signal-to-noise ratio drawback of traditional methods that obtain high frequencies through wavelet compression.

[0141] According to a specific embodiment of the present invention, wavelet elimination is performed in the frequency domain, and the high-frequency spread seismic signal is transformed to the time domain. The specific method includes:

[0142] The frequency of the hour window seismic signal is broadened in the frequency domain and then inverse Fourier transformed to the time domain.

[0143] The frequency of the next hourly window is then overridden until the frequency of the entire seismic signal is overridden.

[0144] Repeat the above two steps for each signal of the entire 3D seismic data to perform frequency upscaling, thereby obtaining high-resolution seismic data.

[0145] According to one specific embodiment of the present invention, the hour window of the Fourier transform is determined based on the time window of the seismic data.

[0146] According to a specific embodiment of the present invention, broadening the frequency of the hour window seismic signal in the frequency domain and performing an inverse Fourier transform to the time domain includes:

[0147] After removing the wavelet from the high-frequency spread domain of the seismic signal, the following formula is obtained:

[0148]

[0149] Actual seismic wavelets are time-varying and space-varying, making it impossible to obtain the true seismic wavelets. In other words, it is impossible to directly obtain the high frequency of the seismic signal by compressing the wavelets. This invention innovatively removes the wavelets in the high-frequency extended frequency domain of the seismic signal, thus eliminating the influence of the wavelets.

[0150]

[0151] If the reflection coefficient of the segment within the longitudinal hour window is uniformly distributed white noise, and the reflection within the hour window is equal before and after scaling, then the following formula applies:

[0152]

[0153] The high-frequency spread of the seismic signal within the nth time window is expressed by the following formula:

[0154]

[0155] in,

[0156] The frequency domain reflection coefficient after scaling;

[0157] This is the frequency domain seismic signal after scaling.

[0158] T p The hour window length for the Fourier transform;

[0159] n = 1, 2, 3... indicates that the hourly window length is T. p The time window number of the captured seismic signal;

[0160] Indicates that the hourly window length is T p The frequency domain reflection coefficient of the nth time window;

[0161] The length of the hour window after scaling is T. p The frequency domain reflection coefficient of the nth time window;

[0162] Indicates that the hourly window length is T p The seismic signal with frequency spread in the frequency domain of the nth time window;

[0163] The length of the hour window after scaling is T. p The nth time window frequency domain seismic signal;

[0164] According to a specific embodiment of the present invention, the frequency extension of the immediately following hour window is performed until the frequency extension of the entire seismic signal is completed. Specifically, the frequency-spread seismic signals in all hour windows are added together to obtain a complete high-frequency spread seismic signal in the frequency domain.

[0165]

[0166] Performing an inverse Fourier transform on the above equation yields the seismic signal after high-frequency time-domain expansion:

[0167] s(t) = F -1 (s) hf (f))

[0168] in,

[0169] N represents the hourly window length as T. p The total number of seismic signals captured.

[0170] s” hf (f) represents the hourly window length T. p A complete seismic signal with high-frequency extension in the frequency domain;

[0171] s(t) is s” hf (f) Time-domain seismic signal after inverse Fourier transform;

[0172] T is the total length of the seismic signal;

[0173] F -1 This represents the inverse Fourier transform.

[0174] According to a specific embodiment of the present invention, determining the reliability of high-frequency extended data using well data and known geological information specifically includes:

[0175] By using well logging acoustic waves and density curves to create synthetic records and comparing them with high-frequency extended seismic data, the fidelity and amplitude of the high-frequency extended seismic data are determined.

[0176] According to a specific embodiment of the present invention, structural interpretation and reservoir prediction using seismic data specifically includes: performing seismic structural interpretation and reservoir prediction on high-frequency extended seismic data.

[0177] According to a specific embodiment of the present invention, seismic tectonic interpretation and reservoir prediction on high-frequency extended seismic data includes seismic tectonic interpretation and reservoir prediction through at least one of layer tracking, fault interpretation, and formation elastic parameter inversion.

[0178] According to a specific embodiment of the present invention, seismic tectonic interpretation and reservoir prediction are performed through stratigraphic tracing, specifically as follows:

[0179] By synthesizing seismic records from acoustic and density data from well logging, the characteristics and location of seismic phase axes corresponding to the stratigraphic layers that need to be traced and interpreted are identified. The phase axes of the stratigraphic layers are then traced in the high-frequency extended seismic signal profile to obtain the spatial distribution characteristics of the layers.

[0180] According to a specific embodiment of the present invention, seismic tectonic interpretation and reservoir prediction through fault interpretation specifically involves:

[0181] Identify whether there is a fault on the seismic phase axis corresponding to the tracked layer. The fault is the reflection feature of the underground fault. Tracking and interpreting the fault feature is called fault interpretation. The interpretation of layer and fault is collectively called structural interpretation.

[0182] According to a specific embodiment of the present invention, seismic tectonic interpretation and reservoir prediction using formation elastic parameters specifically involve:

[0183] The formation reflection coefficient is obtained by using sparse pulse deconvolution on the high-frequency extended seismic signal.

[0184] Then, the reflection coefficient is integrated and combined with the low-frequency components of the low-frequency model of elastic parameters established by the well logging curves to realize the inversion of formation elastic parameters using high-frequency extended seismic signals;

[0185] The values ​​of the elastic parameters corresponding to the reservoir are obtained through well logging interpretation. Based on the formation elastic parameters obtained by inversion, the characteristics and range of the reservoir are delineated using the values ​​of the elastic parameters corresponding to the reservoir.

[0186] The present invention provides a high-frequency seismic signal extension device, employing any of the above-mentioned high-frequency seismic signal extension methods, comprising: a seismic signal acquisition module, a seismic signal frequency extension module, and a high-frequency extension result verification module;

[0187] The seismic signal acquisition module performs the following operations:

[0188] Acquire high-fidelity, amplitude-preserving post-stack seismic data that requires high-frequency extension, and extract seismic signals from the acquired high-fidelity, amplitude-preserving post-stack seismic data;

[0189] The seismic signal frequency extension module performs the following operations:

[0190] Transform the seismic signal into the frequency domain;

[0191] High-frequency extension of seismic signals in the frequency domain;

[0192] Eliminate the wavelet in the frequency domain and transform the high-frequency spread seismic signal to the time domain;

[0193] The high-frequency extension result verification and utilization module performs the following operations:

[0194] Using well data and known geological information, the reliability of high-frequency extended data is determined, and seismic data is used for structural interpretation and reservoir prediction.

[0195] Example 1

[0196] According to a specific embodiment of the present invention, the high-frequency extension method for seismic signals of the present invention will be described in detail below.

[0197] This invention provides a method for high-frequency spread of seismic signals, comprising the following steps:

[0198] Acquire high-fidelity, amplitude-preserving post-stack seismic data that requires high-frequency extension, and extract seismic signals from the acquired high-fidelity, amplitude-preserving post-stack seismic data;

[0199] Transform the seismic signal into the frequency domain;

[0200] High-frequency extension of seismic signals is performed in the frequency domain; before extension, the high-frequency signal is weak, and after extension, the energy is enhanced, allowing for the identification of more subtle details.

[0201] Eliminate the wavelet in the frequency domain and transform the high-frequency spread seismic signal to the time domain;

[0202] Using well data and known geological information, the reliability of high-frequency extended data is determined, and seismic data is used for structural interpretation and reservoir prediction.

[0203] Example 2

[0204] According to a specific embodiment of the present invention, the high-frequency extension method for seismic signals of the present invention will be described in detail below.

[0205] This invention provides a method for high-frequency spread of seismic signals, comprising the following steps:

[0206] Acquire high-fidelity, amplitude-preserving post-stack seismic data that requires high-frequency extension, and extract seismic signals from the acquired high-fidelity, amplitude-preserving post-stack seismic data;

[0207] Transforming seismic signals to the frequency domain includes specific methods such as:

[0208] Time-domain seismic signals are represented by the following formula:

[0209] s(t)=r(t)*w(t)

[0210] The seismic signal transformed to the frequency domain is represented by the following equation:

[0211] s'(f)=F(s(t))=r'(f)·w'(f)

[0212] in,

[0213] s(t) is the time-domain seismic signal;

[0214] s'(f) is the frequency domain seismic signal;

[0215] w(t) is a time-domain seismic wavelet;

[0216] w'(f) is a frequency-domain seismic wavelet;

[0217] r(t) is the time-domain reflection coefficient;

[0218] r'(f) is the frequency domain reflection coefficient;

[0219] F represents the Fourier transform.

[0220] The high-frequency extension of seismic signals in the frequency domain is performed as follows:

[0221] The high-frequency spread of seismic signals in the time domain is expressed by the following formula:

[0222] s hf (t)=r(t)*w(at)

[0223] The transformation of a time-domain high-frequency spread seismic signal to the frequency domain is expressed by the following formula:

[0224]

[0225] in,

[0226] s hf (t) represents the high-frequency spread seismic signal in the time domain;

[0227] s' hf (f) is the seismic signal transformed from the time-domain high-frequency spread seismic signal to the frequency domain;

[0228] w(at) is the time-domain seismic wavelet after scaling transformation;

[0229] This is the frequency domain seismic wavelet after scaling transformation;

[0230] a is the scaling factor, where a > 1.

[0231] Eliminating the wavelet in the frequency domain and transforming the high-frequency spread seismic signal to the time domain, specific methods include:

[0232] The frequency of the hour window seismic signal is broadened in the frequency domain and then inverse Fourier transformed to the time domain.

[0233] The frequency of the next hourly window is then spread until the frequency of the entire seismic signal is spread.

[0234] Repeat the above two steps for each signal of the entire 3D seismic data to perform frequency upscaling, thereby obtaining high-resolution seismic data.

[0235] Using well data and known geological information, the reliability of high-frequency extended data is determined, and seismic data is used for structural interpretation and reservoir prediction.

[0236] Example 3

[0237] According to a specific embodiment of the present invention, the high-frequency extension method for seismic signals of the present invention will be described in detail below.

[0238] This invention provides a method for high-frequency spread of seismic signals, comprising the following steps:

[0239] Acquire high-fidelity, amplitude-preserving post-stack seismic data that requires high-frequency extension, and extract seismic signals from the acquired high-fidelity, amplitude-preserving post-stack seismic data;

[0240] Transforming seismic signals to the frequency domain includes specific methods such as:

[0241] Time-domain seismic signals are represented by the following formula:

[0242] s(t)=r(t)*w(t)

[0243] The seismic signal transformed to the frequency domain is represented by the following equation:

[0244] s'(f)=F(s(t))=r'(f)·w'(f)

[0245] in,

[0246] s(t) is the time-domain seismic signal;

[0247] s'(f) is the frequency domain seismic signal;

[0248] w(t) is a time-domain seismic wavelet;

[0249] w'(f) is a frequency-domain seismic wavelet;

[0250] r(t) is the time-domain reflection coefficient;

[0251] r'(f) is the frequency domain reflection coefficient;

[0252] F represents the Fourier transform.

[0253] The high-frequency extension of seismic signals in the frequency domain is performed as follows:

[0254] The high-frequency spread of seismic signals in the time domain is expressed by the following formula:

[0255] s hf (t)=r(t)*w(at)

[0256] The transformation of a time-domain high-frequency spread seismic signal to the frequency domain is expressed by the following formula:

[0257]

[0258] in,

[0259] s hf (t) represents the high-frequency spread seismic signal in the time domain;

[0260] s' hf (f) is the seismic signal transformed from the time-domain high-frequency spread seismic signal to the frequency domain;

[0261] w(at) is the time-domain seismic wavelet after scaling transformation;

[0262] This is the frequency domain seismic wavelet after scaling transformation;

[0263] a is the scaling factor, where a > 1.

[0264] Eliminating the wavelet in the frequency domain and transforming the high-frequency spread seismic signal to the time domain, specific methods include:

[0265] Based on the time window of the seismic data, the hour window for Fourier transform is determined. The frequency of the hour window seismic signal is broadened in the frequency domain and then inverse Fourier transformed to the time domain. Specifically, this includes:

[0266] Removing the wavelet from the high-frequency spread domain of the seismic signal yields the following formula:

[0267]

[0268] If the reflection coefficient of the segment within the longitudinal hour window is uniformly distributed white noise, and the reflection within the hour window is equal before and after scaling to three decimal places, then the following formula applies:

[0269]

[0270] The high-frequency spread of the seismic signal within the nth time window is expressed by the following formula:

[0271]

[0272] in,

[0273] The frequency domain reflection coefficient after scaling;

[0274] This is the frequency domain seismic signal after scaling.

[0275] T p The hour window length for the Fourier transform;

[0276] n = 1, 2, 3... indicates that the hourly window length is T. p The time window number of the captured seismic signal;

[0277] Indicates that the hourly window length is T p The frequency domain reflection coefficient of the nth time window;

[0278] The length of the hour window after scaling is T. p The frequency domain reflection coefficient of the nth time window;

[0279] Indicates that the hourly window length is T p The seismic signal with frequency spread in the frequency domain of the nth time window;

[0280] The length of the hour window after scaling is T. p The nth time window frequency domain seismic signal;

[0281] The frequency extension of the next hourly window is performed until the entire seismic signal is extended. Specifically, the frequency-spread seismic signals within all hourly windows are summed together to obtain a complete high-frequency spread seismic signal.

[0282]

[0283] Performing an inverse Fourier transform on the above equation yields the seismic signal after high-frequency time-domain expansion:

[0284] s(t) = F -1 (s) hf (f))

[0285] in,

[0286] N represents the hourly window length as T. p The total number of seismic signals captured.

[0287] s” hf (f) represents the hourly window length T. p A complete seismic signal with high-frequency extension in the frequency domain;

[0288] s(t) is s”hf (f) Time-domain seismic signal after inverse Fourier transform;

[0289] T is the total length of the seismic signal;

[0290] F -1 This represents the inverse Fourier transform.

[0291] Repeat the above two steps for each signal of the entire 3D seismic data to perform frequency upscaling, thereby obtaining high-resolution seismic data.

[0292] Using well data and known geological information, the reliability of high-frequency extended data is determined, and seismic data is used for structural interpretation and reservoir prediction.

[0293] Example 4

[0294] According to a specific embodiment of the present invention, the high-frequency extension method for seismic signals of the present invention will be described in detail below.

[0295] This invention provides a method for high-frequency spread of seismic signals, comprising the following steps:

[0296] Acquire high-fidelity, amplitude-preserving post-stack seismic data that requires high-frequency extension, and extract seismic signals from the acquired high-fidelity, amplitude-preserving post-stack seismic data;

[0297] Transforming seismic signals to the frequency domain includes specific methods such as:

[0298] Time-domain seismic signals are represented by the following formula:

[0299] s(t)=r(t)*w(t)

[0300] The seismic signal transformed to the frequency domain is represented by the following equation:

[0301] s'(f)=F(s(t))=r'(f)·w'(f)

[0302] in,

[0303] s(t) is the time-domain seismic signal;

[0304] s'(f) is the frequency domain seismic signal;

[0305] w(t) is a time-domain seismic wavelet;

[0306] w'(f) is a frequency-domain seismic wavelet;

[0307] r(t) is the time-domain reflection coefficient;

[0308] r'(f) is the frequency domain reflection coefficient;

[0309] F represents the Fourier transform.

[0310] The high-frequency extension of seismic signals in the frequency domain is performed as follows:

[0311] The high-frequency spread of seismic signals in the time domain is expressed by the following formula:

[0312] s hf (t)=r(t)*w(at)

[0313] The transformation of a time-domain high-frequency spread seismic signal to the frequency domain is expressed by the following formula:

[0314]

[0315] in,

[0316] s hf (t) represents the high-frequency spread seismic signal in the time domain;

[0317] s' hf (f) is the seismic signal transformed from the time-domain high-frequency spread seismic signal to the frequency domain;

[0318] w(at) is the time-domain seismic wavelet after scaling transformation;

[0319] This is the frequency domain seismic wavelet after scaling transformation;

[0320] a is the scaling factor, where a > 1.

[0321] Eliminating the wavelet in the frequency domain and transforming the high-frequency spread seismic signal to the time domain, specific methods include:

[0322] Based on the time window of the seismic data, the hour window for Fourier transform is determined. The frequency of the hour window seismic signal is broadened in the frequency domain and then inverse Fourier transformed to the time domain. Specifically, this includes:

[0323] Removing the wavelet from the high-frequency spread domain of the seismic signal yields the following formula:

[0324]

[0325] If the reflection coefficient of the segment within the longitudinal hour window is uniformly distributed white noise, and the reflection within the hour window is equal before and after scaling to three decimal places, then the following formula applies:

[0326]

[0327] The high-frequency spread of the seismic signal within the nth time window is expressed by the following formula:

[0328]

[0329] in,

[0330] The frequency domain reflection coefficient after scaling;

[0331] This is the frequency domain seismic signal after scaling.

[0332] T p The hour window length for the Fourier transform;

[0333] n = 1, 2, 3... indicates that the hourly window length is T. p The time window number of the captured seismic signal;

[0334] Indicates that the hourly window length is T p The frequency domain reflection coefficient of the nth time window;

[0335] The length of the hour window after scaling is T. p The frequency domain reflection coefficient of the nth time window;

[0336] Indicates that the hourly window length is T p The seismic signal with frequency spread in the frequency domain of the nth time window;

[0337] The length of the hour window after scaling is T. p The nth time window frequency domain seismic signal;

[0338] The frequency extension of the next hourly window is performed until the entire seismic signal is extended. Specifically, the frequency-spread seismic signals within all hourly windows are summed together to obtain a complete high-frequency spread seismic signal.

[0339]

[0340] Performing an inverse Fourier transform on the above equation yields the seismic signal after high-frequency time-domain expansion:

[0341] s(t) = F -1 (s) hf (f))

[0342] in,

[0343] N represents the hourly window length as T. p The total number of seismic signals captured.

[0344] s” hf (f) represents the hourly window length T. p A complete seismic signal with high-frequency extension in the frequency domain;

[0345] s(t) is s”hf (f) Time-domain seismic signal after inverse Fourier transform;

[0346] T is the total length of the seismic signal;

[0347] F -1 This represents the inverse Fourier transform.

[0348] Repeat the above two steps for each signal of the entire 3D seismic data to perform frequency upscaling, thereby obtaining high-resolution seismic data.

[0349] Using well data and known geological information, the reliability of high-frequency extended data is determined, and seismic data is used for structural interpretation and reservoir prediction. Specifically, determining the reliability of high-frequency extended data using well data and known geological information includes:

[0350] By using well logging sonic waves and density curves to create synthetic records and comparing them with high-frequency extended seismic data, the fidelity and amplitude of the high-frequency extended seismic data are determined.

[0351] Structural interpretation and reservoir prediction using seismic data specifically include: performing seismic structural interpretation and reservoir prediction on high-frequency extended seismic data, including at least one of the following: horizon tracking, fault interpretation, and formation elastic parameter inversion.

[0352] Seismic tectonic interpretation and reservoir prediction are performed through horizon tracing, specifically as follows:

[0353] By synthesizing seismic records from acoustic and density data from well logging, the characteristics and location of seismic phase axes corresponding to the stratigraphic layers that need to be traced and interpreted are identified. The phase axes of the stratigraphic layers are then traced in the high-frequency extended seismic signal profile to obtain the spatial distribution characteristics of the layers.

[0354] Seismic tectonic interpretation and reservoir prediction through fault interpretation specifically involve:

[0355] Identify whether there is a fault on the seismic phase axis corresponding to the tracked layer. The fault is the reflection feature of the underground fault. Tracking and interpreting the fault feature is called fault interpretation. The interpretation of layer and fault is collectively called structural interpretation.

[0356] Seismic tectonic interpretation and reservoir prediction using formation elastic parameters specifically involve:

[0357] The formation reflection coefficient is obtained by using sparse pulse deconvolution on the high-frequency spread seismic signal.

[0358] Then, the reflection coefficient is integrated and combined with the low-frequency components of the low-frequency model of elastic parameters established by the well logging curves to realize the inversion of formation elastic parameters using high-frequency extended seismic signals;

[0359] The values ​​of the elastic parameters corresponding to the reservoir are obtained through well logging interpretation. Based on the formation elastic parameters obtained by inversion, the characteristics and range of the reservoir are delineated using the values ​​of the elastic parameters corresponding to the reservoir.

[0360] Example 5

[0361] According to a specific embodiment of the present invention, the high-frequency seismic signal extension device of the present invention will be described in detail below.

[0362] The present invention provides a high-frequency seismic signal extension device, employing any of the above-mentioned high-frequency seismic signal extension methods, comprising: a seismic signal acquisition module, a seismic signal frequency extension module, and a high-frequency extension result verification module;

[0363] The seismic signal acquisition module performs the following operations:

[0364] Acquire high-fidelity, amplitude-preserving post-stack seismic data that requires high-frequency extension, and extract seismic signals from the acquired high-fidelity, amplitude-preserving post-stack seismic data;

[0365] The seismic signal frequency extension module performs the following operations:

[0366] Transform the seismic signal into the frequency domain;

[0367] High-frequency extension of seismic signals in the frequency domain;

[0368] Eliminate the wavelet in the frequency domain and transform the high-frequency spread seismic signal to the time domain;

[0369] The high-frequency extension result verification and utilization module performs the following operations:

[0370] Using well data and known geological information, the reliability of high-frequency extended data is determined, and seismic data is used for structural interpretation and reservoir prediction.

[0371] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention are included within the scope of protection of the present invention.

Claims

1. A method for high-frequency spread of seismic signals, characterized in that, Includes the following steps: Acquire high-fidelity, amplitude-preserving post-stack seismic data that requires high-frequency extension, and extract seismic signals from the acquired high-fidelity, amplitude-preserving post-stack seismic data; Transform the seismic signal into the frequency domain; Specific methods for transforming seismic signals to the frequency domain include: Time-domain seismic signals are represented by the following formula: The seismic signal transformed to the frequency domain is represented by the following equation: in, s ( t () represents a time-domain seismic signal; This is a frequency domain seismic signal; w ( t () represents the time-domain seismic wavelet; For frequency domain seismic wavelets; r ( t () represents the time-domain reflection coefficient; The frequency domain reflection coefficient; F Indicates Fourier transform; High-frequency spread of seismic signals in the frequency domain; the specific method for high-frequency spread of seismic signals in the frequency domain is as follows: The high-frequency spread of seismic signals in the time domain is expressed by the following formula: The transformation of a time-domain high-frequency spread seismic signal to the frequency domain is expressed by the following formula: in, This refers to a seismic signal with high-frequency extension in the time domain. This refers to the transformation of a time-domain high-frequency extended seismic signal into a frequency-domain seismic signal; This is the time-domain seismic wavelet after scale transformation; This is the frequency domain seismic wavelet after scaling transformation; The scaling factor. ; Eliminate the wavelet in the frequency domain and transform the high-frequency spread seismic signal to the time domain; Using well data and known geological information, the reliability of high-frequency extended data is determined, and seismic data is used for structural interpretation and reservoir prediction.

2. The high-frequency spread method for seismic signals according to claim 1, characterized in that, Eliminating the wavelet in the frequency domain and transforming the high-frequency spread seismic signal to the time domain, specific methods include: The frequency of the hour window seismic signal is broadened in the frequency domain and then inverse Fourier transformed to the time domain. The frequency of the next hourly window is then overridden until the frequency of the entire seismic signal is overridden. Repeat the above two steps for each signal of the entire 3D seismic data to perform frequency extension, thereby obtaining high-resolution seismic data.

3. The high-frequency spread method for seismic signals according to claim 2, characterized in that, The hour window for Fourier transform is determined based on the time window of the seismic data.

4. The high-frequency spread method for seismic signals according to claim 3, characterized in that, The frequency broadening of the hour window seismic signal in the frequency domain and its inverse Fourier transform to the time domain includes: Removing the wavelet from the high-frequency spread domain of the seismic signal yields the following formula: If the reflection coefficient of the segment within the longitudinal hour window is uniformly distributed white noise, and the reflection within the hour window is equal before and after scaling to three decimal places, then the following formula applies: The first n The high-frequency spread of the seismic signal within a time window is expressed by the following formula: in, The frequency domain reflection coefficient after scaling; This is the frequency domain seismic signal after scaling. T p The hour window length for the Fourier transform; n =1,2,3... indicates a window length of... T p The time window number of the captured seismic signal; Indicates the hourly window length is T p The frequency domain reflection coefficient of the nth time window; The length of the hour window after scaling is... T p The frequency domain reflection coefficient of the nth time window; Indicates the hourly window length is T p The n Seismic signals with frequency spread in the frequency domain of a time window; The length of the hour window after scaling is... T p The n Seismic signals in the frequency domain of a time window.

5. The high-frequency spread method for seismic signals according to claim 4, characterized in that, The frequency extension of the next adjacent hourly window is performed until the entire seismic signal is extended. Specifically, the frequency-spread seismic signals within all hourly windows are summed together to obtain a complete high-frequency spread seismic signal in the frequency domain. Performing an inverse Fourier transform on the above equation yields the seismic signal after high-frequency time-domain expansion: in, N Indicates the length of the hour window as T p The total number of seismic signals captured. ; The hourly window length is T p A complete seismic signal with high-frequency extension in the frequency domain; for Time-domain seismic signal after inverse Fourier transform; T This represents the total length of the seismic signal. F -1 This represents the inverse Fourier transform.

6. The high-frequency spread method for seismic signals according to claim 5, characterized in that, Determining the reliability of high-frequency extended data using well data and known geological information specifically includes: By using well logging sonic waves and density curves to create synthetic records and comparing them with high-frequency extended seismic data, the fidelity and amplitude of the high-frequency extended seismic data are determined.

7. The high-frequency spread method for seismic signals according to claim 6, characterized in that, Using seismic data for structural interpretation and reservoir prediction specifically includes performing seismic structural interpretation and reservoir prediction on high-frequency extended seismic data.

8. The high-frequency spread method for seismic signals according to claim 6, characterized in that, Seismic tectonic interpretation and reservoir prediction on high-frequency extended seismic data includes seismic tectonic interpretation and reservoir prediction through at least one of the following: horizon tracking, fault interpretation, and formation elastic parameter inversion.

9. The high-frequency spread method for seismic signals according to claim 8, characterized in that, Seismic tectonic interpretation and reservoir prediction are performed through horizon tracing, specifically as follows: By synthesizing seismic records from acoustic and density data from well logging, the characteristics and location of seismic phase axes corresponding to the stratigraphic layers that need to be traced and interpreted are identified. The phase axes of the stratigraphic layers are then traced in the high-frequency extended seismic signal profile to obtain the spatial distribution characteristics of the layers.

10. The high-frequency spread method for seismic signals according to claim 9, characterized in that, Seismic tectonic interpretation and reservoir prediction through fault interpretation specifically involve: Identify whether there is a fault on the seismic phase axis corresponding to the tracked layer. The fault is the reflection feature of the underground fault. Tracking and interpreting the fault feature is called fault interpretation. The interpretation of layer and fault is collectively called structural interpretation.

11. The high-frequency spread method for seismic signals according to claim 9, characterized in that, Seismic tectonic interpretation and reservoir prediction using formation elastic parameters specifically involve: The formation reflection coefficient is obtained by using sparse pulse deconvolution on the high-frequency extended seismic signal. Then, the reflection coefficient is integrated and combined with the low-frequency components of the low-frequency model of elastic parameters established by the well logging curves to realize the inversion of formation elastic parameters using high-frequency extended seismic signals; The values ​​of the elastic parameters corresponding to the reservoir are obtained through well logging interpretation. Based on the formation elastic parameters obtained by inversion, the characteristics and range of the reservoir are delineated using the values ​​of the elastic parameters corresponding to the reservoir.

12. A high-frequency seismic signal extension device, characterized in that, The high-frequency extension method for seismic signals according to any one of claims 1-11 includes: a seismic signal acquisition module, a seismic signal frequency extension module, and a high-frequency extension result verification module; The seismic signal acquisition module performs the following operations: Acquire high-fidelity, amplitude-preserving post-stack seismic data that requires high-frequency extension, and extract seismic signals from the acquired high-fidelity, amplitude-preserving post-stack seismic data; The seismic signal frequency extension module performs the following operations: Transform the seismic signal into the frequency domain; High-frequency extension of seismic signals in the frequency domain; Eliminate the wavelet in the frequency domain and transform the high-frequency spread seismic signal to the time domain; The high-frequency extension result verification and utilization module performs the following operations: Using well data and known geological information, the reliability of high-frequency extended data is determined, and seismic data is used for structural interpretation and reservoir prediction.

Images