A base-stretching frequency-modulation synchronous extraction seismic reservoir prediction method

Abstract

The application discloses a base stretching and frequency modulation synchronization extraction seismic reservoir prediction method, which comprises the following steps: S1, inputting original two-dimensional seismic profile signal s(x;t) to be analyzed; S2, decomposing the original two-dimensional seismic profile signal s(x;t) by using base stretching adaptive frequency modulation transformation (SBCT) to obtain time-frequency transformation result SBCT(x;f,t); S3, estimating the instantaneous frequency of the xth seismic signal at each time-frequency position according to the time-frequency transformation result SBCT(x;f,t); S4, constructing a synchronization extraction operator with the instantaneous frequency of the signal as the curve center on the time-frequency domain according to the synchronization extraction principle, and using the synchronization extraction operator to extract the original time-frequency spectrum energy to the curve to obtain new time-frequency coefficients and eliminate the ambiguous time-frequency energy; S5, taking the modulus of the time-frequency coefficients to obtain the time-frequency spectrum after base stretching and frequency modulation synchronization extraction transformation; and S6, calculating the main frequency range of the original seismic signal by using Fourier transformation to determine the high-frequency value and the low-frequency value, extracting the spectrum corresponding to the high-frequency value and the low-frequency value from the time-frequency spectrum obtained in S5 to obtain two common frequency profiles, and determining whether the reservoir exists in the seismic profile by comparing the attenuation of the seismic signals in the two common frequency profiles. The application can significantly improve the time-frequency energy focusing property and improve the identification precision of the seismic reservoir.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of signal processing, and particularly relates to a base scaling frequency modulation synchronous extraction seismic reservoir prediction method. BACKGROUND

[0002] Time-frequency analysis is an important means of seismic reservoir prediction, which can effectively characterize the non-stationary characteristics of seismic signals and reveal the relationship between signal frequency and time. As an effective means of signal processing, time-frequency analysis clearly describes the close relationship between signal frequency and time, so it is widely used in seismic reservoir identification. Many scholars have realized oil and gas reservoir identification by using the "absorption and attenuation" law of the instantaneous spectral energy of different frequency slices of geological bodies. In recent years, in order to study complex and variable non-stationary seismic signals and improve the characterization accuracy of signals, scholars have successively proposed a variety of time-frequency analysis methods with high time-frequency resolution, but most of the time-frequency characterization methods cannot process complex multi-component signals without prior conditions.

[0003] The present application proposes a novel chirplet transform method - base scaling adaptive chirplet transform (SBCT) on the basis of chirplet transform (CT), which matches multi-component signals by constructing basis functions. Compared with existing methods, the time-frequency characterization obtained by SBCT method can achieve higher time-frequency energy aggregation. The time-frequency characterization result obtained by synchronous extraction transform (SET) method after energy redistribution of the time-frequency spectrum is more aggregated, and the resolution is greatly improved. Through the above analysis, we know that SET and SBCT have their own advantages in the field of signal processing, so they are jointly analyzed and discussed.

[0004] The basic idea of the base scaling frequency modulation synchronous extraction seismic reservoir prediction method is to obtain the time-frequency characterization result of the signal by SBCT, and then rearrange the time-frequency energy to the real instantaneous frequency ridge of the signal to realize the high aggregation of the time-frequency energy of the signal. The real instantaneous frequency of the signal is estimated in the base scaling adaptive chirplet transform (SBCT) time-frequency domain, the time-frequency energy related to the time-frequency characteristics of the signal is extracted from the original time-frequency spectrum by using the frequency fixed point, and the ambiguous time-frequency energy is removed, so as to accurately depict the time-frequency energy distribution information of the signal and positively promote the accuracy of time-frequency analysis. SUMMARY

[0005] In view of the above deficiencies in the prior art, the present application provides a base scaling frequency modulation synchronous extraction seismic reservoir prediction method. The present application can improve the focusing of the energy distribution of the time-frequency characterization of the signal, improve the identification accuracy of the seismic reservoir, and well reconstruct the signal.

[0006] To achieve the above-mentioned objectives, the technical solution adopted by this invention is as follows: a method for seismic reservoir prediction by synchronous extraction of basic stretching frequency modulation, comprising the following steps:

[0007] S1. Input the original two-dimensional seismic profile signal s(x;t) to be analyzed;

[0008] S2. Perform a basis scaling adaptive frequency modulation transform on the two-dimensional seismic profile signal s(x;t) input from S1 to obtain the time-frequency transformation result SBCT(x;f,t);

[0009] S3. Based on the time-frequency transformation results obtained above, estimate the instantaneous frequency of the x-th seismic signal at each time-frequency location using SBCT(x; f, t);

[0010] S4. Based on the principle of synchronous extraction, a synchronous extraction operator with the instantaneous frequency of the signal as the center of the curve is constructed in the time-frequency domain to extract the original time-frequency spectrum energy and obtain new time-frequency coefficients, thereby eliminating the ambiguous time-frequency energy.

[0011] S5. Take the modulus of the time-frequency coefficients obtained above to obtain the time spectrum after the base-scaling frequency modulation synchronous extraction transformation.

[0012] S6. Calculate the dominant frequency range of the original seismic signal using Fourier transform. Set 85% of the maximum frequency value within the dominant frequency range as high-frequency values ​​and the minimum frequency value as low-frequency values. Then, extract the spectrum corresponding to the high-frequency values ​​from the time spectrum obtained in S5 to obtain the co-frequency profile of the high-frequency values; extract the spectrum corresponding to the low-frequency values ​​from the time spectrum obtained in S5 to obtain the co-frequency profile of the low-frequency values. Determine whether a reservoir exists in the seismic profile by comparing the attenuation of the seismic signal in the two co-frequency profiles.

[0013] Preferably, the time-frequency transformation result of the basis-scaling frequency-modulated adaptive transform of the original two-dimensional seismic profile signal s(x;t) in step S2 is as follows:

[0014]

[0015] Where t represents time, f represents the frequency center, and h(x; τ) represents the Gaussian function. Let τ represent the phase function, where τ is the time variable. for:

[0016]

[0017] Where k = 1, 2, ..., n, and n is the phase function. The order of (a1, a2, ..., a) n ) is the adaptive parameter of the phase function, which can be determined according to kurtosis theory, i.e.:

[0018]

[0019] where argmax(·) denotes the argument set for which the function takes its maximum value.

[0020] As a preference, the instantaneous frequency of the signal at each time-frequency position (t, f) is estimated from the obtained time-frequency transform result SBCT(x; f, t)

[0021]

[0022] where, denotes the partial derivative of the window function h(x; t) with respect to t, denotes the SBCT result under the window function h(x; t). denotes the SBCT result under the function t k h(x; t).

[0023] As a preference, the specific method for obtaining the new time-frequency coefficients in the step S4 is: according to the synchronous extraction principle, a synchronous extraction operator SEO(x; f, t) with the instantaneous frequency of the signal as the curve center is constructed on the time-frequency domain to extract the original time-frequency spectrum energy onto the curve to obtain the new time-frequency coefficients T(x; f, t), and the ambiguous time-frequency energy is removed, and the synchronous extraction operator SEO(x; f, t) satisfies:

[0024]

[0025] where δ(·) is the unit impulse function, and the new time-frequency coefficients T(x; f, t) are:

[0026] T(x; f, t) = SBCT(x; f, t) SEO(x; f, t) (6)

[0027] As a preference, the step S5 can perform inverse transformation on T(x; f, t) in equation (6) to reconstruct the seismic signal s(x; t) by using the following formula:

[0028]

[0029] where,

[0030] The base stretching frequency modulation synchronous extraction seismic reservoir prediction method provided by the application has the beneficial effects that:

[0031] The application takes SBCT algorithm as a breakthrough point, considers the problems of unclear identification and poor aggregation in SET multi-component non-stationary nonlinear complex signals, and proposes a base scaling frequency modulation synchronous extraction algorithm. First, the time-frequency representation result of the signal is obtained through SBCT, then the time-frequency energy is rearranged to the real instantaneous frequency ridge line of the signal to realize high aggregation of the time-frequency energy of the signal; the real instantaneous frequency of the signal is estimated in the SBCT time-frequency domain, the time-frequency energy related to the time-frequency characteristics of the signal is extracted from the original time-frequency spectrum by using the frequency fixed point, and the fuzzy time-frequency energy is removed, so that the time-frequency energy distribution information of the signal is described with high precision, and the precision of time-frequency analysis is improved. The application shows good effect in reconstruction ability and calculation efficiency, can significantly improve the time-frequency energy focusing, and improve the identification precision of the seismic reservoir.

[0032] The idea of the application is:

[0033] Firstly, input the original two-dimensional seismic profile signal s(x;t) to be analyzed;

[0034] Secondly, the base scaling adaptive frequency modulation transformation is performed on the input signal s(x;t), and the time-frequency transformation result SBCT(x;f,t) is obtained by the following calculation method:

[0035]

[0036] Wherein, t represents time, f represents frequency center, h(x;tau) belongs to L 2 (R) represents a Gaussian function, A phase function, tau is a time variable;

[0037] Thirdly, the instantaneous frequency of the signal at each time-frequency position (t,f) is estimated according to the obtained time-frequency transformation result SBCT(x;f,t)

[0038] Fourthly, according to the synchronous extraction principle, a synchronous extraction operator SEO(x;f,t) is constructed on the time-frequency domain, which takes the signal instantaneous frequency curve as the center, is used to extract the original time-frequency spectrum energy to the curve to obtain new time-frequency coefficients T(x;f,t), and remove the fuzzy time-frequency energy;

[0039] Fifthly, the time-frequency coefficients obtained in the fourth step are taken as a module to obtain the time-frequency spectrum after base scaling frequency modulation synchronous extraction.

[0040] Sixth, the main frequency range of the original seismic signal is calculated by using Fourier transform, 85% of the maximum frequency value in the main frequency range is set as the high frequency value, the minimum frequency value in the main frequency range is set as the low frequency value, then the frequency spectrum corresponding to the high frequency value is extracted from the time-frequency spectrum obtained from S5 to obtain the common frequency profile of the high frequency value; the frequency spectrum corresponding to the low frequency value is extracted from the time-frequency spectrum obtained from S5 to obtain the common frequency profile of the low frequency value. Whether the reservoir exists in the seismic profile is determined by comparing the attenuation of the seismic signal in the two common frequency profiles.

[0041] The working principle of the present application is as follows: collecting the original two-dimensional seismic profile signal s(x;t) to be analyzed; using the base stretching adaptive frequency modulation transform (SBCT) to decompose the original two-dimensional seismic profile signal s(x;t) to obtain the time-frequency spectrum SBCT(x;f,t) of the base stretching frequency modulation adaptive transform of the original two-dimensional seismic profile signal s(x;t); estimating the instantaneous frequency of the signal at each time-frequency position according to the time-frequency transform result SBCT(x;f,t); constructing a synchronous extraction operator with the instantaneous frequency of the signal as the center of the curve on the time-frequency domain according to the synchronous extraction principle to extract the original time-frequency spectrum energy to obtain new time-frequency coefficients and eliminate the ambiguous time-frequency energy; taking the modulus of the above time-frequency coefficients to obtain the time-frequency spectrum after the base stretching frequency modulation synchronous extraction transform. The present application can significantly improve the time-frequency energy focusing and improve the identification accuracy of the seismic reservoir.

[0042] The present application takes the SBCT algorithm as the starting point, combines the problems of unclear identification and poor aggregation in the SET multi-component non-stationary nonlinear complex signal, and proposes a base stretching frequency modulation synchronous extraction algorithm. First, the time-frequency representation result of the signal is obtained by SBCT, then the time-frequency energy is rearranged on the real instantaneous frequency ridge line of the signal to realize high aggregation of the time-frequency energy of the signal; the real instantaneous frequency of the signal is estimated in the SBCT time-frequency domain, the time-frequency energy related to the time-frequency characteristics of the signal is extracted from the original time-frequency spectrum by using the frequency fixed point, and the ambiguous time-frequency energy is eliminated, so as to accurately depict the time-frequency energy distribution information of the signal and positively promote the accuracy of time-frequency analysis. The present application has good effects in reconstruction ability and calculation efficiency, can significantly improve the time-frequency energy focusing and improve the identification accuracy of the seismic reservoir. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 The flow chart of the present application is shown in the figure;

[0044] Figure 2 The profile of the Sichuan Basin Zhongjiang gas field in the western part of Central South China is shown in the figure

[0045] Figure 3 The low frequency value common frequency profile of the Zhongjiang gas field processed by the method of the present application is shown in the figure

[0046] Figure 4 High-frequency value common frequency profile of Zhongjiang gas field processed by the method of the application DETAILED DESCRIPTION

[0047] The application will be further described below with reference to the drawings.

[0048] Example 1: see Figure 1 A base stretch frequency modulation synchronous extraction seismic reservoir prediction method, comprising the following steps:

[0049] S1, inputting an original two-dimensional seismic profile signal s(x;t) to be analyzed;

[0050] S2, decomposing the original two-dimensional seismic profile signal s(x;t) by using a base stretch frequency modulation adaptive transform to obtain a base stretch frequency modulation adaptive transform result SBCT(x;f,t) of the original two-dimensional seismic profile signal s(x;t);

[0051] S3, estimating an instantaneous frequency of the signal at each time-frequency position according to the time-frequency transform result SBCT(x;f,t);

[0052] S4, constructing a synchronous extraction operator with the instantaneous frequency of the signal as a curve center on the time-frequency domain according to a synchronous extraction principle to "extract" a new time-frequency coefficient from the original time-frequency spectrum energy and eliminate the ambiguous time-frequency energy;

[0053] S5, taking a modulus of the time-frequency coefficient to obtain a time-frequency spectrum after base stretch frequency modulation synchronous extraction transform.

[0054] S6, calculating a main frequency range of the original seismic signal by using Fourier transform, setting 85% of a frequency maximum value in the main frequency range as a high frequency value and setting a frequency minimum value in the main frequency range as a low frequency value, then extracting a spectrum corresponding to the high frequency value from the time-frequency spectrum obtained in S5 to obtain a high frequency value common frequency profile, and extracting a spectrum corresponding to the low frequency value from the time-frequency spectrum obtained in S5 to obtain a low frequency value common frequency profile. Whether a reservoir exists in the seismic profile is determined by comparing the attenuation of the seismic signal in the two common frequency profiles.

[0055] As a preferred, the time-frequency transform result of the base stretch frequency modulation adaptive transform of the original two-dimensional seismic profile signal s(x;t) in step S2 is:

[0056]

[0057] wherein, t represents time, f represents a frequency center, h(x;τ) represents a Gaussian function, φ(t) represents a phase function, and τ is a time variable;

[0058] The phase function is:

[0059]

[0060] where k = 1, 2,..., n, n is the order of the phase function (a1, a2,..., an) are adaptive parameters of the phase function, which can be determined according to the kurtosis theory, i.e.: n

[0061]

[0062] where argmax(·) represents the parameter set when the function takes the maximum value.

[0063] As a preferred, the step S3 of estimating the instantaneous frequency of the signal at each time-frequency position (t, f) according to the time-frequency spectrum phase information is:

[0064]

[0065] where represents the partial derivative of the window function h(x; t) with respect to t, represents the SBCT result under the window function . represents the SBCT result under the function t k h(x; t).

[0066] As a preferred, the specific method of obtaining the new time-frequency coefficient in the step S4 is: according to the synchronous extraction principle, a synchronous extraction operator SEO(x; f, t) with the instantaneous frequency of the signal as the center curve is constructed on the time-frequency domain to extract the original time-frequency spectrum energy to the curve to obtain the new time-frequency coefficient T(x; f, t), and the ambiguous time-frequency energy is removed, and the synchronous extraction operator SEO(x; f, t) satisfies:

[0067]

[0068] where δ(·) is a unit impulse function, and the new time-frequency coefficient T(x; f, t) is:

[0069] T(x; f, t) = SBCT(x; f, t) SEO(x; f, t) (6)

[0070] As a preferred, the step S5 can perform inverse transformation on T(x; f, t) in equation (6) to reconstruct the seismic signal s(x; t)

[0071]

[0072] where

[0073] Referring to​Figures 1 to 4 We take a seismic profile as an example, and the two-dimensional seismic profile is shown in the figure. Figure 2 Figures 3-4 The low-frequency value common frequency profile and the high-frequency value common frequency profile are extracted from the time-frequency spectrum obtained by performing the base-stretching self-adapting frequency modulation transform on the seismic profile. In the figure, the horizontal coordinate represents the seismic trace number, the vertical coordinate represents the time, and the right color bar represents the energy value. The embodiment proves that the result obtained by the method has higher time-frequency resolution and more concentrated energy, and the identification accuracy of the seismic reservoir is improved.

[0074] The above only describes the preferred embodiments of the present application and is not used to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.​

Claims

1. A base-stretch frequency-modulated synchronous extraction seismic reservoir prediction method, characterized in that, The method comprises the following steps: S1, inputting a seismic signal s(x;t) to be analyzed, wherein x represents a seismic trace, and t represents time; S2, performing base-stretching adaptive frequency modulation transformation on the seismic signal s(x;t) input in S1 to obtain a time-frequency transformation result SBCT(x;f,t), and the calculation method is as follows: where t denotes time, f e R denotes a frequency center, h(x; τ) denotes a Gaussian function, denotes a phase function, and τ is a time variable; S3, estimating the instantaneous frequency of the xth seismic signal at each time-frequency location (t, f) from the time-frequency transformed result SBCT(x; f, t) obtained in S2 S4, according to the principle of synchronous extraction, a synchronous extraction operator is constructed in time-frequency domain, which is centered on the signal instantaneous frequency curve For extracting the original time-frequency spectrum energy, a new time-frequency coefficient T(x;f, t) is obtained, and the ambiguous time-frequency energy is eliminated. S5, taking the modulus of the time-frequency coefficient obtained in S5 to obtain a time-frequency spectrum T(x;f,t) after base-stretching frequency modulation synchronous extraction transformation; S6, calculating a main frequency range of the seismic signal s(x;t) by using Fourier transformation, setting 85% of a maximum frequency value in the main frequency range as a high frequency value, setting a minimum frequency value in the main frequency range as a low frequency value, then extracting a spectrum corresponding to the high frequency value from the time-frequency spectrum obtained in S5 to obtain a common frequency profile of the high frequency value, extracting a spectrum corresponding to the low frequency value from the time-frequency spectrum obtained in S5 to obtain a common frequency profile of the low frequency value, and determining whether a reservoir exists in the seismic profile by comparing attenuation of the seismic signal in the two common frequency profiles.

2. The base-stretch frequency-modulated synchronous extraction transform seismic reservoir prediction method according to claim 1, characterized in that, The time-frequency spectrum of the base-stretching adaptive frequency modulation transformation of the seismic signal s(x;t) in the step S2 is as follows: where t denotes time, f denotes the frequency center, h(x; τ) denotes a Gaussian function, denotes a phase function, τ is a time variable, and the phase function is: where k = 1, 2, …, n, n is the order of the phase function , (a1, a2, …, a n ) are adaptive parameters of the phase function , which can be determined according to the kurtosis theory, i.e. Wherein, argmax(·) represents a parameter set when a function takes a maximum value.

3. The base-stretch frequency-modulated synchronous extraction seismic reservoir prediction method of claim 1, wherein, The calculation of the instantaneous frequency in said step S3 comprises estimating the instantaneous frequency of the signal at each time-frequency position (t,f) from the time-frequency transform result SBCT(x;f,t) obtained in S2 wherein denotes the partial derivative of the window function h(x; t) with respect to t, denotes the SBCT result computed under the window function h(x; t). denotes the SBCT result computed under the function t k h(x; t).

4. The base-stretch frequency-modulated synchronous extraction seismic reservoir prediction method of claim 1, wherein, The specific method for obtaining the new time-frequency coefficient T(x;f,t) in step S4 is: according to the synchronous extraction principle, constructing a synchronous extraction operator with the signal instantaneous frequency as the curve center in the time-frequency domain The synchronous extraction operator extracts the original time-frequency spectrum energy to obtain the new time-frequency coefficient, and eliminates the ambiguous time-frequency energy Satisfies: Wherein, δ(·) is a unit impulse function, and the new time-frequency coefficient T(x;f,t) is as follows:

5. The base-stretch frequency-modulated synchronous extraction seismic reservoir prediction method of claim 4, wherein, The step S5 can perform inverse transformation on T(x;f,t) in formula (6) to reconstruct the seismic signal s(x;t) by using the following formula: wherein,

Images