A method for detecting interfaces of thin interbeds in tight oil based on synchronous extrusion generalized S transform

Abstract

The present invention provides a method for detecting a thin interlayer interface of tight oil based on synchronous extrusion generalized S transform, comprising: step 1: performing generalized S transform on seismic signals to obtain a generalized S transform time spectrum; step 2: performing synchronous extrusion processing on the generalized S transform time spectrum to obtain a synchronous extrusion generalized S transform time spectrum; step 3: performing threshold filtering and Gaussian smoothing processing on the synchronous extrusion generalized S transform time spectrum; step 4: extracting a single frequency curve from the processed synchronous extrusion generalized S transform time spectrum, and searching for the time corresponding to all peak amplitudes on the single frequency curve, thus obtaining a detection result of a thin interlayer interface of tight oil. The method for detecting a thin interlayer interface of tight oil based on synchronous extrusion generalized S transform has higher time-frequency resolution and stronger anti-noise capability, and thus can obtain a detection result of a thin interlayer interface of tight oil with higher resolution.

Description

Technical Field

[0001] The invention relates to the technical field of seismic signal processing, and in particular to a method for detecting a tight oil thin interlayer interface based on synchronous extrusion generalized S transform. Background Art

[0002] With the continuous development of oil and gas exploration in my country, the exploration targets have gradually shifted to complex and concealed reservoirs such as tight oil and gas reservoirs and thin interbeds. However, the resolution of seismic data of relevant concealed reservoirs is generally low, which is difficult to meet the requirements of fine structural interpretation and fine reservoir prediction. The existing conventional high-resolution processing methods often fail to achieve ideal results when processing seismic data of complex structural blocks such as tight oil and gas reservoirs. The current high-resolution processing methods suitable for complex structural seismic data are imperfect, which seriously restricts the development of tight oil and gas and thin reservoir exploration.

[0003] As an important means of analyzing non-stationary signals, time-frequency analysis plays an important role in high-resolution processing of seismic data, and predecessors have conducted a lot of research in this area. In 2003, Gao Jinghuai et al. proposed a four-parameter generalized S transform and studied its application in thin interlayer detection; in 2005, Wan Tao et al. studied a three-parameter wavelet transform and applied it to the characterization of the internal structure of thin interlayers; in 2013, Zhang Yang studied the application of the S transform frequency division method in the analysis of tight oil reservoirs; in 2018, Han Chongyang studied the application of wavelet transform and S transform in high-resolution interpretation of tight oil reservoirs; in 2019, Xia Yaliang et al. applied the generalized S transform multi-frequency interpretation technology to the analysis of thin layers. Although the traditional time-frequency analysis method has achieved certain results in the high-resolution processing of tight oil and gas reservoirs and thin reservoir seismic data, there are still defects that limit the application effect.

[0004] First, due to the defects of different time-frequency analysis methods, the time-frequency resolution cannot meet the requirements, and the accuracy of the interface identification results of tight oil thin interlayers is not high. Short-time Fourier transform uses a fixed window function. When this method is used to analyze seismic signals, the resolution of different frequency components in the time-frequency spectrum is the same, which does not meet the characteristics of seismic signals. Wavelet transform uses a variable window function, but its result is in the time-scale domain, and its physical specific meaning is unclear when used for seismic signal analysis. S transform uses a variable-scale Gaussian window function, which can achieve higher frequency resolution when analyzing the low-frequency components of the signal and higher time resolution when analyzing the high-frequency components of the signal, and its result is in the time-frequency domain, which is conducive to the analysis of seismic signals; generalized S transform introduces multiple parameters on the basis of S transform to adjust the speed of change of Gaussian window function with frequency, so as to obtain a time-frequency spectrum that is more in line with actual needs. However, the time-frequency spectrum of generalized S transform still has the phenomenon of energy divergence, so when used for the detection of tight oil thin interlayer interfaces, the resolution of thin layers is still limited.

[0005] On the other hand, when using traditional time-frequency analysis methods to identify the interface of tight oil thin interlayers, the data must have a high signal-to-noise ratio. The time-frequency resolution of traditional time-frequency analysis methods is limited, and the performance characteristics of thin interlayers in their time-frequency spectra are more susceptible to noise interference, which significantly reduces the effect of identification and even leads to false interfaces, resulting in erroneous conclusions.

[0006] In the Chinese patent application with application number: CN202010014657.9, a seismic acoustic impedance inversion method based on time-varying wavelet and curvelet transform constraints is involved, which belongs to the field of geophysical exploration technology, and specifically includes the following steps: performing an improved generalized S transform on two-dimensional zero-offset seismic data, and extracting time-varying wavelets from the time-frequency spectrum using spectral simulation technology; according to the sparsity of the reflection coefficient, performing L1 norm-constrained reflection coefficient inversion based on the time-varying wavelet to obtain high-precision reflection coefficient data; according to the continuous formation model, taking the reflection coefficient inversion result as input data, constructing and solving the multi-channel acoustic impedance inversion objective function based on curvelet transform constraints, and obtaining a laterally enhanced relative acoustic impedance profile.

[0007] In the Chinese patent application with application number: CN201811502989.0, a method for simulating the seismic wave propagation matrix of an arbitrary complex medium is involved, and the method includes the following steps: 1) dividing the non-uniform geological model of the horizontal surface into a number of horizontal thin layers with a thickness of Δz; 2) performing Fourier transform on the generalized Lippmann-Schwinger integral equation in each horizontal thin layer to derive the reflection / transmission matrix; 3) utilizing the boundary continuity conditions between the horizontal thin layers to derive the generalized reflection / transmission matrix through the reflection / transmission matrix; 4) using the inter-layer iterative relationship, using the generalized reflection / transmission matrix to solve the seismic wave field in the frequency domain; 5) using the inverse Fourier transform to convert the seismic wave field to the time domain to achieve the desired goal.

[0008] In the Chinese patent application with application number: CN201710910942.7, a method and system for predicting a seismic reflection interface is involved, including: 1) taking one of the seismic data as input data to establish a seismic signal function; 2) extracting wavelets from the data selected in step 1) to establish a wavelet function; 3) based on the seismic signal function and the wavelet function, obtaining a time-frequency analysis function of the seismic signal, and detecting whether each time sample point is a seismic reflection point within a set frequency band; 4) taking the next seismic data for detection until all seismic data are detected to obtain a seismic reflection interface.

[0009] In summary, the current time-frequency analysis method has the problem of limited time-frequency resolution and susceptibility to noise interference when identifying the interface of tight oil thin interlayers. Therefore, it is necessary to propose a time-frequency domain tight oil thin interlayer interface detection method with high time-frequency resolution and strong noise resistance. To this end, we invented a new tight oil thin interlayer interface detection method based on synchronous extrusion generalized S transform. Summary of the invention

[0010] The purpose of the present invention is to provide a method for detecting thin interbed interfaces of tight oil based on synchronous extrusion generalized S transform, which solves the problems of low time-frequency resolution and poor noise resistance in the existing time-frequency domain thin interbed interface detection methods of tight oil.

[0011] The object of the present invention can be achieved by the following technical measures: a method for detecting the interface of a thin interlayer of tight oil based on synchronous extrusion generalized S transform, the method for detecting the interface of a thin interlayer of tight oil based on synchronous extrusion generalized S transform comprising:

[0012] Step 1: Perform generalized S transform on the seismic signal to obtain the generalized S transform time-frequency spectrum;

[0013] Step 2: Perform synchronous squeezing processing on the generalized S transform time spectrum to obtain a synchronous squeezed generalized S transform time spectrum;

[0014] Step 3: Threshold filtering and Gaussian smoothing are performed on the spectrum of the synchronous squeezed generalized S transform;

[0015] Step 4: Extract the single-frequency curve from the processed synchronous compression generalized S-transform time spectrum, and find the time corresponding to all peak amplitudes on the single-frequency curve to obtain the detection result of the interface of the tight oil thin interlayer.

[0016] The purpose of the present invention can also be achieved by the following technical measures:

[0017] In step 1, appropriate parameters λ and p are selected for the seismic signal h(t) to be analyzed, and the generalized S transform is obtained to obtain the generalized S transform time-frequency spectrum:

[0018]

[0019] Among them, S(τ,f) is the time-frequency spectrum of the generalized S transform; h(t) is the seismic signal to be analyzed; λ and p are the adjustment parameters of the Gaussian window function, which are used to adjust the speed at which the window function of the generalized S transform changes with frequency; τ is the time sampling point; f is the frequency sampling point; t is the time series corresponding to the signal h(t) to be analyzed.

[0020] In step 2, the generalized S transform time-frequency spectrum obtained in step 1 is subjected to synchronous squeezing processing to obtain a synchronous squeezed generalized S transform time-frequency spectrum; first, the instantaneous frequency value of each point in the time-frequency spectrum is calculated; then, based on the obtained instantaneous frequency value, the amplitude value of each point in the generalized S transform time-frequency spectrum is placed at its true instantaneous frequency and superimposed.

[0021] In step 2, the formula for calculating the instantaneous frequency value of each point in the time spectrum is:

[0022]

[0023] Where: w(τ,f) is the instantaneous frequency; τ is the time of the point to be analyzed; f is the frequency of the point to be analyzed, and S(τ,f) is the generalized S-transform time-frequency spectrum.

[0024] In step 2, the amplitude value of each point in the generalized S-transform spectrum S(τ,f) is placed at its true instantaneous frequency and the formula for superposition is:

[0025]

[0026] Where: is the synchronous squeezed generalized S transform time spectrum; L f is the frequency w in the spectrum during the generalized S transform l Half the length of the frequency interval centered at w l is the frequency axis of the spectrum during synchronous squeezing generalized S transform; Δw is the frequency axis L f The frequency interval ofk is the frequency axis sampling interval of the spectrum during generalized S-transformation, f k is the frequency axis of the spectrum during generalized S-transformation.

[0027] In step 3, the synchronous squeezed generalized S transform time spectrum obtained in step 2 is subjected to threshold filtering denoising and Gaussian smoothing processing; first, the synchronous squeezed generalized S transform time spectrum T obtained in step 2 is calculated. S (τ,w l ) is used to calculate the noise threshold according to the standard deviation of the noise; then, the synchronous extrusion time-frequency spectrum is subjected to hard threshold denoising according to the calculated noise threshold; finally, Gaussian smoothing is performed on the synchronous extrusion generalized S transform time-frequency spectrum after threshold filtering and denoising to suppress the amplitude spectrum anomaly caused by non-thin interlayers, and at the same time, the interference of noise can be further suppressed.

[0028] In step 3, the synchronous squeezed generalized S transform time spectrum T obtained in step 2 is calculated S (τ,w l ) is calculated as:

[0029]

[0030] Where: ση is the standard deviation of the noise; T S (τ,w l ) is the spectrum of the synchronously squeezed generalized S transform; median(·) indicates the calculated median.

[0031] In step 3, the formula for calculating the noise threshold γ based on the standard deviation of the noise is:

[0032]

[0033] Where n is the number of sampling points of the signal h(t).

[0034] In step 3, the formula for hard threshold denoising of the synchronous squeezed spectrum is:

[0035]

[0036] Where: abs(·) means finding the amplitude of the time-frequency spectrum.

[0037] In step 3, the formula for Gaussian smoothing of the synchronous squeezed generalized S transform time spectrum after threshold filtering and denoising is:

[0038]

[0039] Where: T(τ, w l ) is the time-frequency spectrum after Gaussian smoothing; T S (τ,w l) is the time spectrum before Gaussian smoothing; is the control parameter of Gaussian smoothing function; x and y are the variables of Gaussian smoothing function, τ is the time of the point to be analyzed; w l is the frequency axis of the spectrum during the synchronously squeezed generalized S-transform.

[0040] In step 4, in the processed synchronous extrusion generalized S-transform time-frequency spectrum, the single-frequency curve T(τ, w0) at the frequency w0 is extracted, and all τ values ​​that satisfy it are found, that is, the time corresponding to each peak value of the single-frequency curve; all the τ values ​​obtained are the detection results of the thin interlayer interface of tight oil.

[0041] In step 4, in the processed synchronous squeezed generalized S transform time spectrum, extract the single frequency curve T(τ, w0) at the frequency w0, and find the formula for all τ values ​​that satisfy:

[0042]

[0043] Where: T(τ, w0) is the time-frequency spectrum when the time is τ and the frequency is w0; Δτ is the time sampling interval; abs(·) represents the amplitude calculation.

[0044] Compared with the prior art, the method for detecting the interface of thin interbedded tight oil layers based on synchronous extrusion generalized S-transformation in the present invention has the following advantages and positive effects:

[0045] The present invention adopts the synchronous extrusion generalized S transform with more accurate time-frequency results and higher resolution to detect the thin interlayer interface of tight oil, which effectively solves the problems of inaccurate detection results and low resolution caused by low time-frequency resolution in the prior art; at the same time, the new synchronous extrusion generalized S transform adopted has stronger noise resistance than the prior art, and can provide a more accurate time-frequency spectrum under the influence of noise. Therefore, when the present invention is used for thin interlayer interface detection of tight oil seismic data with low signal-to-noise ratio, more accurate and reliable results can be obtained. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a flow chart of a specific embodiment of the method for detecting the interface of tight oil thin interlayer based on synchronous extrusion generalized S transform of the present invention;

[0047] Figure 2 It is a schematic diagram of the time and coefficient size of 6 thin interlayer interfaces simulated in a specific embodiment of the present invention;

[0048] Figure 3 To use the Ricker wavelet with a main frequency of 20Hz Figure 2 Schematic diagram of the noisy thin interbed seismic signal obtained by convolving the reflection coefficient in and adding random noise;

[0049] Figure 4 It is a schematic diagram of the generalized S transform time spectrum after threshold filtering and Gaussian smoothing in a specific embodiment of the present invention;

[0050] Figure 5 It is a schematic diagram of the spectrum of the synchronous squeezed generalized S transform after threshold filtering and Gaussian smoothing in a specific embodiment of the present invention;

[0051] Figure 6 It is a schematic diagram of a single frequency curve extracted from a generalized S-transform spectrum in a specific embodiment of the present invention;

[0052] Figure 7 It is a schematic diagram of a single frequency curve extracted from a spectrum during synchronously squeezed generalized S-transform in a specific embodiment of the present invention;

[0053] Figure 8 A comparison diagram of the generalized S-transform thin interlayer interface detection result and the real reflection interface position in a specific embodiment of the present invention;

[0054] Fig. 9 A comparison diagram of the interface detection result of the tight oil thin interlayer by synchronous extrusion generalized S transformation and the actual interface position in a specific embodiment of the present invention;

[0055] Fig.10 It is a schematic diagram of an actual seismic section of a certain tight oil thin interbed development area in a specific embodiment of the present invention;

[0056] Fig.11 In a specific embodiment of the present invention, a generalized S transform method is used to transform Fig.10 The schematic diagram of the result of detecting the interface of thin interbeds of tight oil using the seismic section shown;

[0057] Fig.12 In a specific embodiment of the present invention, a method of synchronously squeezing a generalized S transform is used to Fig.10 The schematic diagram of the result of detecting the interface of thin interbeds of tight oil using the seismic section shown;

[0058] Fig.13 It is a schematic diagram of an actual seismic section of a certain tight oil thin interbed development area in a specific embodiment of the present invention;

[0059] Fig.14 In a specific embodiment of the present invention, a generalized S transform method is used to transform Fig.13 The schematic diagram of the result of detecting the interface of thin interbeds of tight oil using the seismic section shown;

[0060] Fig.15 In a specific embodiment of the present invention, a method of synchronously squeezing a generalized S transform is used to Fig.13The schematic diagram of the result of detecting the interface of thin interbeds of tight oil using the seismic section shown;

[0061] Fig.16 It is a schematic diagram of an actual seismic section of a certain tight oil thin interbed development area in a specific embodiment of the present invention;

[0062] Fig.17 In a specific embodiment of the present invention, a generalized S transform method is used to transform Fig.16 The schematic diagram of the result of detecting the interface of thin interbeds of tight oil using the seismic section shown;

[0063] Fig.18 In a specific embodiment of the present invention, a method of synchronously squeezing a generalized S transform is used to Fig.16 The schematic diagram shown is a result of detecting the interface of thin interbeds of tight oil using a seismic profile. DETAILED DESCRIPTION

[0064] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0065] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations and / or combinations thereof.

[0066] like Figure 1 As shown, Figure 1 The flowchart of the method for detecting the thin interlayer reflection interface of tight oil based on synchronous extrusion generalized S transform of the present invention is as follows:

[0067] S 1: Select appropriate parameters λ and p for the seismic signal h(t) to be analyzed and perform generalized S transform to obtain the generalized S transform time-frequency spectrum:

[0068]

[0069] Among them, S(τ,f) is the time-frequency spectrum of the generalized S transform; h(t) is the seismic signal to be analyzed; λ and p are the adjustment parameters of the Gaussian window function, which are used to adjust the speed at which the window function of the generalized S transform changes with frequency; τ is the time sampling point; f is the frequency sampling point; t is the time series corresponding to the signal h(t) to be analyzed.

[0070] S2: Perform synchronous squeezing processing on the generalized S-transform time-frequency spectrum obtained in the previous step to obtain a synchronous squeezed generalized S-transform time-frequency spectrum;

[0071] First, calculate the instantaneous frequency value of each point in the time-frequency spectrum:

[0072]

[0073] Where: w(τ,f) is the instantaneous frequency; τ is the time of the point to be analyzed; is the frequency of the point to be analyzed, and S(τ,f) is the generalized S-transform time-frequency spectrum.

[0074] Then, according to the instantaneous frequency value obtained, the amplitude value of each point in the generalized S-transform time spectrum S(τ,f) is placed at its true instantaneous frequency and superimposed:

[0075]

[0076] Where: T S (τ,w l ) is the synchronous squeezed generalized S transform time spectrum; L f is the frequency w in the spectrum during the generalized S transform l Half the length of the frequency interval centered at w l is the frequency axis of the spectrum during synchronous squeezing generalized S transform; Δw is the frequency axis w l The frequency interval of k is the frequency axis sampling interval of the spectrum during generalized S-transform.

[0077] S3: Perform threshold filtering denoising and Gaussian smoothing on the synchronous squeezed generalized S transform time spectrum obtained in the previous step;

[0078] First, calculate the synchronous squeezed generalized S transform time spectrum T obtained in the previous step S (τ,w l ) is the standard deviation of the noise:

[0079]

[0080] Where: η is the standard deviation of the noise; is the spectrum of the synchronous squeezed generalized S transform; median(·) indicates the calculated median.

[0081] The noise threshold is calculated based on the standard deviation of the noise:

[0082]

[0083] Where n is the number of sampling points of the signal h(t).

[0084] Then, according to the noise threshold calculated in the previous step, hard threshold denoising is performed on the synchronous extrusion spectrum:

[0085]

[0086] Where: abs(·) means finding the amplitude of the time-frequency spectrum.

[0087] Finally, Gaussian smoothing is performed on the synchronous squeezed generalized S transform time spectrum after threshold filtering and denoising to suppress the amplitude spectrum anomaly caused by non-thin interlayers and further suppress the interference of noise:

[0088]

[0089] Where: T(τ, w l ) is the time-frequency spectrum after Gaussian smoothing; T S (τ,w l ) is the time-frequency spectrum before Gaussian smoothing; σ is the control parameter of the Gaussian smoothing function; x and y are the variables of the Gaussian smoothing function.

[0090] S4: Extract the single frequency curve T(τ, w0) at the frequency w0 from the processed synchronous compression generalized S transform time spectrum obtained in the previous step, and find all the τ values ​​that satisfy it, that is, the time corresponding to each peak of the single frequency curve. All the τ values ​​obtained are the detection results of the thin interlayer reflection interface of tight oil.

[0091]

[0092] Where: T(τ, w0) is the time-frequency spectrum when the time is τ and the frequency is f; Δτ is the time sampling interval; abs(·) represents the amplitude calculation.

[0093] Embodiment 1:

[0094] In a specific embodiment of the present invention, the method for detecting the interface of tight oil thin interlayers based on synchronous extrusion generalized S transform includes the following steps:

[0095] like Figure 2 The model shown here has 6 thin interlayer interfaces, and the reflection coefficients are 1, -1, 1, 1, -1 and 1, respectively, and the corresponding time positions are 65 milliseconds, 75 milliseconds, 85 milliseconds, 95 milliseconds, 105 milliseconds and 115 milliseconds. The above reflection coefficient sequence is convolved with a Ricker wavelet with a main frequency of 20 Hz, and then random noise is added to obtain the following Figure 3 The noisy seismic signal shown. According to the definition of thin interlayer, when the layer thickness is less than one quarter of the wavelength, it is considered to be a thin interlayer. Therefore, when the main frequency of the Ricker wavelet is 20Hz, it is considered to be a thin layer when the time thickness is less than 12.5 milliseconds. Figure 2 The time thickness of the middle layer is 10 milliseconds, which meets the condition of thin interlayer. Figure 3It can be seen from the noisy seismic signal in that it is difficult to accurately identify the positions of the six reflection interfaces from the simulated seismic signal due to the thin layer thickness. The noisy signal is subjected to generalized S transform and synchronous squeezed generalized S transform with the same parameters, and the obtained time-frequency spectrum is subjected to threshold filtering and Gaussian smoothing respectively. The processed generalized S transform time-frequency spectrum and synchronous squeezed generalized S transform time-frequency spectrum are shown as follows: Figure 4 and Figure 5 As shown. It can be seen that under the same transformation parameters, the synchronous extrusion generalized S transform time-frequency spectrum used in the present invention has a higher time-frequency resolution than the traditional method, and the reflection of the thin interlayer on the time-frequency spectrum is clearer. The single frequency curve at the frequency of 45Hz is extracted from the time-frequency spectrum obtained by the above two methods. The results are shown in Figure 6 and Figure 7 As shown in Figure 2. By comparison, it can be seen that the single-frequency curve extracted by synchronous extrusion generalized S transform has a higher time resolution and a clearer description of the interface of the tight oil thin interlayer. The detection results of the tight oil thin interlayer interface are obtained by extracting the time corresponding to the peak of the above two single-frequency curves. The comparison diagram of the recognition effect of the two methods and the position of the real reflection interface is shown in Figure 2. Figure 8 and Fig. 9 shown. Figure 8 and 9 In the figure, positive values ​​indicate the generalized S transform detection results, and negative values ​​indicate the true reflection interface position. Comparative analysis shows that the traditional generalized S transform method only detected three reflection interface positions, with the results being 73 milliseconds, 90 milliseconds, and 115 milliseconds, respectively. It was not able to fully detect the positions of all thin interlayer reflection interfaces, and the results differed greatly from the true positions. The synchronous extrusion generalized S transform method detected the positions of all six thin-layer reflection interfaces, with the results being 65 milliseconds, 73 milliseconds, 85 milliseconds, 95 milliseconds, 107 milliseconds, and 116 milliseconds, respectively. The error of its detection result was significantly smaller than that of the traditional generalized S transform. The results of the model data test show that, compared with the existing methods, the invented method can significantly improve the accuracy of the detection results of the thin interlayer reflection interface of tight oil.

[0096] Embodiment 2:

[0097] like Fig.10 The figure shows an actual seismic profile of a certain area with thin interbedded tight oil layers, with a total of 110 traces and a time range of 625 milliseconds to 825 milliseconds. Analysis shows that the seismic data has a low resolution and weak energy, making it difficult to effectively distinguish the position of the reflection interface, which is not conducive to the subsequent interpretation work. The traditional method based on generalized S transform and the method based on synchronous extrusion generalized S transform proposed in this invention are used to detect the reflection interface of thin interbedded tight oil layers. The detection results are shown as follows: Fig.11 and Fig.12As shown in the figure, it can be seen from the comparison that some reflection interfaces that have not been identified by the traditional method have been effectively identified in the method proposed in the present invention, and the resolution of the identification result has been significantly improved, which shows that the method can be effectively applied to the detection of reflection interfaces of thin interlayers of tight oil in actual seismic data.

[0098] Embodiment 3:

[0099] Fig.13 This is another typical seismic profile of the same tight oil thin interbed development area as Example 2. The profile has a total of 88 traces with a time range of 100 milliseconds to 350 milliseconds. Fig.14 The detection result of the reflection interface of the thin interlayer of tight oil obtained by using the generalized S transform method is shown in Figure 2. Fig.15 The detection result of the interface of the thin interbed of tight oil obtained by using the synchronous extrusion generalized S transform method. By comparison, it can be seen that the recognition result of the method proposed in the present invention has obviously higher resolution, which is more conducive to the analysis of the work area.

[0100] Embodiment 4:

[0101] Fig.16 This is a typical seismic profile of another area with thin tight oil interbeds. The profile has a total of 110 traces with a time range of 1000 milliseconds to 1350 milliseconds. Fig.17 and Fig.18 The results of identification of tight oil thin interlayer reflection interface obtained by using the generalized S transform of the prior art and the new synchronous extrusion generalized S transform proposed by the present invention are shown respectively. By comparison, it can be seen that compared with the prior art, the new technology proposed by the present invention can identify the tight oil thin interlayer interface that the prior art has not identified, and the identification result is more accurate, which can effectively improve the accuracy of the seismic profile interpretation.

[0102] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions recorded in the aforementioned embodiments or replace some of the technical features therein with equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0103] Except for the technical features described in the specification, all other technical features are known technologies to those skilled in the art.

Claims

1. A method for detecting the interface of tight oil thin interbeds based on synchronous extrusion generalized S transform, characterized in that: The method for detecting the interface of tight oil thin interbeds based on synchronous extrusion generalized S transform includes: Step 1: Perform generalized S transform on the seismic signal to obtain the generalized S transform time-frequency spectrum; Step 2: Perform synchronous squeezing processing on the generalized S transform time spectrum to obtain a synchronous squeezed generalized S transform time spectrum; Step 3: Threshold filtering and Gaussian smoothing are performed on the spectrum of the synchronous squeezed generalized S transform; Step 4: extract the single-frequency curve from the processed synchronous compression generalized S-transform time spectrum, and find the time corresponding to all peak amplitudes on the single-frequency curve, so as to obtain the detection result of the interface of the tight oil thin interlayer; In step 3, the synchronous squeezed generalized S transform time spectrum obtained in step 2 is subjected to threshold filtering denoising and Gaussian smoothing processing; first, the synchronous squeezed generalized S transform time spectrum T obtained in step 2 is calculated. S (τ,w l ) is used to calculate the noise threshold according to the standard deviation of the noise; then, the synchronous extrusion time spectrum is subjected to hard threshold denoising according to the calculated noise threshold; finally, Gaussian smoothing is performed on the synchronous extrusion generalized S transform time spectrum after threshold filtering denoising to suppress the amplitude spectrum anomaly caused by non-thin interlayers, and at the same time, the interference of noise can be further suppressed; In step 3, the synchronous squeezed generalized S transform time spectrum T obtained in step 2 is calculated S (τ,w l ) is calculated as: Where: η is the standard deviation of the noise; T S (τ,w l ) is the spectrum of synchronous squeezed generalized S transform; median(·) indicates the calculated median; In step 3, the formula for calculating the noise threshold γ based on the standard deviation of the noise is: Where n is the number of sampling points of the signal h(t); In step 3, the formula for hard threshold denoising of the synchronous squeezed spectrum is: Where: abs(·) means finding the amplitude of the time-frequency spectrum; In step 3, the formula for Gaussian smoothing of the synchronous squeezed generalized S transform time spectrum after threshold filtering and denoising is: Where: T(τ,w l ) is the time-frequency spectrum after Gaussian smoothing; T S (τ,w l ) is the time-frequency spectrum before Gaussian smoothing; σ is the control parameter of Gaussian smoothing function; x and y are the variables of Gaussian smoothing function, τ is the time of the point to be analyzed; w l It is the frequency axis of the spectrum during synchronous squeezing generalized S transform; In step 4, in the processed synchronous extrusion generalized S-transform time spectrum, a single frequency curve T(τ,w0) at a frequency of w0 is extracted, and all τ values ​​that satisfy it are found, that is, the time corresponding to each peak value of the single frequency curve; all the τ values ​​obtained are the detection results of the interface of the thin interlayer of tight oil; In step 4, in the processed synchronous squeezed generalized S transform time spectrum, extract the single frequency curve T(τ,w0) at the frequency w0, and find the formula for all τ values ​​that satisfy: Where: T(τ,w0) is the time-frequency spectrum when the time is τ and the frequency is w0; Δτ is the time sampling interval; abs(·) represents the amplitude calculation.

2. The method for detecting the interface of tight oil thin interbeds based on synchronous extrusion generalized S transform according to claim 1, characterized in that: In step 1, appropriate parameters λ and p are selected for the seismic signal h(t) to be analyzed, and the generalized S transform is obtained to obtain the generalized S transform time-frequency spectrum: Among them, S(τ,f) is the time-frequency spectrum of the generalized S transform; h(t) is the seismic signal to be analyzed; λ and p are the adjustment parameters of the Gaussian window function, which are used to adjust the speed at which the window function of the generalized S transform changes with frequency; τ is the time sampling point; f is the frequency sampling point; t is the time series corresponding to the signal h(t) to be analyzed.

3. The method for detecting the interface of tight oil thin interbeds based on synchronous extrusion generalized S transform according to claim 1, characterized in that: In step 2, the generalized S transform time-frequency spectrum obtained in step 1 is subjected to synchronous squeezing processing to obtain a synchronous squeezed generalized S transform time-frequency spectrum; first, the instantaneous frequency value of each point in the time-frequency spectrum is calculated; then, based on the obtained instantaneous frequency value, the amplitude value of each point in the generalized S transform time-frequency spectrum is placed at its true instantaneous frequency and superimposed.

4. The method for detecting the interface of tight oil thin interbeds based on synchronous extrusion generalized S transform according to claim 3 is characterized in that: In step 2, the formula for calculating the instantaneous frequency value of each point in the time spectrum is: Where: w(τ,f) is the instantaneous frequency; τ is the time of the point to be analyzed; f is the frequency of the point to be analyzed, and S(τ,f) is the generalized S-transform time-frequency spectrum.

5. The method for detecting the interface of tight oil thin interbeds based on synchronous extrusion generalized S transform according to claim 4 is characterized in that: In step 2, the amplitude value of each point in the generalized S-transform spectrum S(τ,f) is placed at its true instantaneous frequency and the formula for superposition is: Where: T S (τ,w l ) is the synchronous squeezed generalized S transform time spectrum; L f is the frequency w in the spectrum during the generalized S transform l Half the length of the frequency interval centered at w l is the frequency axis of the spectrum during synchronous squeezing generalized S transform; Δw is the frequency axis w l The frequency interval of k is the frequency axis sampling interval of the spectrum during generalized S transform; f k is the frequency axis of the spectrum during generalized S-transformation.

Images