A gather amplitude-preserved frequency-raising detuning correction method based on complex domain wiener filter

Abstract

The present application relates to the technical field of geophysical exploration of oil and gas resources, and more particularly to a gather amplitude-preserved frequency-raised detuning correction method based on complex domain Wiener filtering. The gather amplitude-preserved frequency-raised detuning correction method based on complex domain Wiener filtering comprises the following steps: constructing an analytical signal of a pre-stack gather complex domain and obtaining a maximum instantaneous amplitude of the pre-stack gather; selecting a target trace from the pre-stack gather; performing Wiener filtering on the remaining traces in the pre-stack gather except the target trace according to the target trace and by obtaining a Wiener filter operator of the analytical signal; obtaining a maximum instantaneous amplitude of the pre-stack gather after Wiener filtering, and determining a proportional coefficient of the maximum instantaneous amplitude before Wiener filtering and the maximum instantaneous amplitude after Wiener filtering; and obtaining a required pre-stack gather by performing AVO amplitude compensation on the pre-stack gather after Wiener filtering. The present application overcomes the problem of missing seismic data when processing the pre-stack gather.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of geophysical exploration of oil and gas resources, and more particularly to a gather amplitude-preserved frequency-raised detuning correction method based on complex domain Wiener filtering. BACKGROUND

[0002] With the deepening of oil and gas exploration, the exploration of conventional oil and gas resources has entered the final stage, and strong air variation non-homogeneous thin oil and gas reservoirs such as sandstone and gravel bodies and river facies thin reservoirs have gradually become one of the new exploration targets. For the problem of characterization and identification of thin oil and gas reservoirs, the inversion description accuracy is subject to the quality of prestack gather data. Due to the large amount of data in the process of seismic data processing, it is difficult to give priority to the fine processing of all gathers, and the gathers after NMO correction are easily affected by the tuning effect in the shallow layer and large offset, and the reflection event waveform is severely stretched and distorted. The inversion based on the prestack gather with tuning effect will seriously affect the accuracy of the inversion result.

[0003] The existing prestack gather detuning processing method mainly has the following problems: 1) the traditional prestack gather detuning processing method relies on gather distortion cutting, which causes the loss of shallow layer and large offset seismic data, the loss of shallow layer geological information, and the artificial reduction of the number of shallow layer coverages, directly losing a large amount of effective information and reducing the subsequent inversion accuracy; 2) the traditional prestack gather detuning processing method does not consider the influence of frequency on gather data, and the frequency of near and far reflection events decreases from near to far, which is not conducive to the fine characterization of thin interbeds.

[0004] The prior art discloses a fracture reservoir rapid prediction method based on prestack seismic data. According to the quality of prestack seismic data, the seismic data is processed by adopting the combination optimization of denoising, flattening and cutting. On the gather after optimization processing, the amplitude values of each trace at a certain time point are fitted, the fitting values of each trace at the time point are subtracted from the average value of the near-gather amplitude values of each trace in the uniform medium, and the difference values of each trace are obtained. The standard deviation thereof is taken as an abnormal value, which can effectively represent the development strength of the fracture reservoir. With the sampling rate as the step, the time point is moved to obtain the abnormal value data body of the entire seismic data. The disclosed technical scheme takes the fracture difference as the basis for fracture reservoir prediction, and comprehensively considers the comprehensive influence of fractures and fluids. The problem of fracture reservoir prediction of wide and narrow azimuth seismic data is quickly and effectively solved. However, the disclosed technical scheme processes the seismic data by cutting, which loses a large amount of effective information and reduces the subsequent inversion accuracy. SUMMARY

[0005] The present application is a kind of trace amplitude-preserving frequency-raising detuning correction method based on complex domain Wiener filtering to overcome the problem of missing seismic data in the prior art when detuning prestack traces.

[0006] To solve the above technical problems, the technical scheme adopted by the present application is as follows: a kind of trace amplitude-preserving frequency-raising detuning correction method based on complex domain Wiener filtering, comprising the following:

[0007] S1: construct the analytic signal of the prestack trace complex domain and calculate the maximum instantaneous amplitude of the prestack trace;

[0008] S2: select a target trace from the prestack trace in step S1;

[0009] S3: according to the target trace in step S2, the Wiener filter operator of the analytic signal in step S1 is used to perform Wiener filtering on the remaining traces in the prestack trace except the target trace;

[0010] S4: calculate the maximum instantaneous amplitude of the prestack trace after Wiener filtering in step S3, and determine the proportional coefficient of the maximum instantaneous amplitude before and after Wiener filtering of the prestack trace;

[0011] S5: AVO amplitude compensation is performed on the prestack trace after Wiener filtering in step S4 to obtain the required prestack trace.

[0012] In the technical scheme, by constructing an analytic signal, selecting a target trace and performing Wiener filtering on the remaining traces in the prestack trace, and then performing amplitude compensation on the prestack trace after Wiener filtering, the required prestack trace is obtained. The technical scheme avoids the drawbacks of the traditional trace distortion removal method, such as missing seismic data in shallow layers and large offset, losing shallow geological information, and artificially reducing the number of shallow layers.

[0013] Preferably, the construction of the analytic signal of the trace complex domain in step S1 comprises the following:

[0014] Assuming that the total number of prestack traces is n, the prestack trace is:

[0015] S(t) = [s1(t), s2(t), s3(t), …, s n (t)] (1);

[0016] Performing Hilbert transform on any trace in the prestack trace:

[0017]

[0018] The analytic signal expression of the trace is:

[0019]

[0020] The analytic signal in the complex field of the pre-stack trace set is:

[0021] Z(t)=[z1(t),z2(t),z3(t),…,z n (t)] (4);

[0022] In the formula, S(t) is a function of the pre-stack gather with respect to time, s n (t) is a function of time for the nth seismic trace. Let s(t) be a function of time after undergoing the Hilbert transform, z(t) be a function of time for each analytic signal, and Z(t) be a function of time for the analytic signal of the pre-stack gather. Here, t is time, τ is a mathematical variable, and i is the imaginary unit. In this technical solution, the essence of the analytic signal is to transform a narrowband real signal into a complex signal, and the Hilbert transform can generate a unique complex signal for a narrowband real signal.

[0023] Preferably, the step S1 of determining the maximum instantaneous amplitude of the pre-stack gather includes the following:

[0024] The sequential amplitude at any point in each analytical signal is:

[0025]

[0026] The instantaneous amplitude of the pre-stack set is:

[0027] A(t)=[a1(t),a2(t),a3(t),…,a n (t)] (6);

[0028] In the formula, a(t) is the time-varying amplitude at any point in each analytical signal, and s(t) is a function of time for each seismic trace. Let s(t) be a function of time after the Hilbert transform, A(t) be the instantaneous amplitude of the pre-stack gather, and a be the time function. n (t) represents the sequential amplitude at any point in the nth analytical signal.

[0029] Preferably, the Wiener filter operator for obtaining the analytic signal in step S3 includes the following:

[0030] Suppose Z(t) = [z1(t), z2(t), z3(t), ..., z n The analytical signal z1(t) of the first seismic trace within the range [(t)] is taken as the target trace, and the analytical signal z2(t) of the second seismic trace is taken as the trace to be matched and processed. The filtering expectation of the analytical signal z1(t) of the first seismic trace as the analytical signal z2(t) of the second seismic trace is continuously approximated. Then the actual expected output of z2(t) is... It can be represented as:

[0031]

[0032] desired output

[0033] where h(t) is a filter operator, is expressed as an extension of the Wiener filter, t is time.

[0034] Preferably, the desired output may also be expressed in integral form:

[0035]

[0036] The desired error of the analytic signal z2(t) of the second seismic trace is:

[0037]

[0038] The total energy mean square error of the analytic signal z2(t) of the second seismic trace is:

[0039]

[0040] where h(t) is a filter operator, τ is a mathematical variable, z1(t) is the analytic signal of the first seismic trace, z2(t) is the analytic signal of the second seismic trace, is expressed as an extension of the Wiener filter.

[0041] Preferably, according to equations (9) to (11), the total energy mean square error of the analytic signal z2(t) of the second seismic trace can also be expressed as:

[0042]

[0043] changing the integral variable:

[0044]

[0045] The total energy mean square error of the analytic signal z2(t) of the second seismic trace is calculated according to equations (12) and (13):

[0046]

[0047] where ε(t) is the desired error of the analytic signal z2(t) of the second seismic trace, E[ε 2 (t)] is the total energy mean square error of the analytic signal z2(t) of the second seismic trace, is represented as a continuation of a Wiener filter, h(t) is a filter operator, t is time, z1(t) is an analytic signal of a first seismic trace, z2(t) is an analytic signal of a second seismic trace, τ is a mathematical variable, v is a mathematical variable, is represented as an autocorrelation of z1(t), is represented as a cross-correlation of z1(t) and z2(t), is represented as an autocorrelation of z2(t).

[0048] Preferably, a partial derivative of a total energy mean square error of the analytic signal z2(t) of the second seismic trace is calculated and the partial derivative is set to zero:

[0049]

[0050] Meanwhile, a mathematical expression of the matched filter operator can be represented as a linear equation group:

[0051]

[0052] A matrix form of the equation (16) is:

[0053]

[0054] In the equation, h z2 = [h(0), h(1), …, h(m)] is an operator matrix of each trace in a matched pre-stack trace set;

[0055] Then, an operator matrix of the entire pre-stack trace set is:

[0056]

[0057] In the equation, ε(t) is an expected error of the analytic signal z2(t) of the second seismic trace, E[ε 2 (t)] is a total energy mean square error of the analytic signal z2(t) of the second seismic trace, is represented as a continuation of a Wiener filter, h(t) is a filter operator, t is time, z1(t) is an analytic signal of a first seismic trace, z2(t) is an analytic signal of a second seismic trace, τ is a mathematical variable, v is a mathematical variable, is represented as an autocorrelation of z1(t), is represented as a cross-correlation of z1(t) and z2(t), is represented as an autocorrelation of z2(t), m represents a number of rows or columns of a matrix, and H represents an operator matrix of the entire pre-stack trace set.

[0058] Preferably, the step S4 of obtaining a maximum instantaneous amplitude of the pre-stack trace set after the Wiener filtering includes the following:

[0059] An analytic signal of the pre-stack trace set after the Wiener filtering in a complex domain is:

[0060]

[0061]

[0062]

[0063]

[0064]

[0065]

[0066]

[0067]

[0068]

[0069]

[0070]

[0071]

[0072] Wiener

[0073]

[0074] ​​​​​​​​​​​​​​​​​

[0075] Preferably, the target trace in the step S2 is a seismic trace with high frequency in the pre-stack gather. In the technical solution, the generation of the tuning effect in the pre-stack gather is mainly caused by the change of the time thickness and the seismic frequency. The large offset is prone to produce waveform stretching distortion due to the decrease of the time thickness and the change of the frequency component. The target trace in the pre-stack gather is recorded at the near offset, the reflection layer is obvious, and the information of the target layer is well reflected. The generation of the tuning effect is mainly caused by the change of the time thickness and the seismic frequency. The large offset is prone to produce waveform stretching distortion due to the decrease of the time thickness and the change of the frequency component. In order to correct and even eliminate the influence, the target trace is preferably selected at the near offset.

[0076] Compared with the prior art, the beneficial effects of the present application are:

[0077] In the present application, the required pre-stack gather is obtained by constructing the analytic signal, selecting the target trace, performing Wiener filtering on the rest of the pre-stack gather, and then performing amplitude compensation on the filtered pre-stack gather. The technical solution avoids the disadvantages of the traditional distortion cutting method, such as the loss of shallow layer seismic data, the loss of shallow layer geological information, and the artificial reduction of the number of coverages of the shallow layer. The present application maintains the consistency of the main frequency and the spectral bandwidth of the wavelet at different offsets, weakens and even eliminates the tuning effect of the large offset caused by the decrease of the time thickness, greatly improves the seismic resolution of the gather, and enhances the ability to describe thin reservoirs. BRIEF DESCRIPTION OF DRAWINGS

[0078] Figure 1 is the flow chart of the amplitude-preserving frequency-raising detuning correction method of the gather based on the complex domain Wiener filtering of the present application;

[0079] Figure 2 is the wave shape display of the wave gather recording of the double-layer single-interface model and its acoustic wave forward modeling in embodiment 1 after cutting off the direct wave;

[0080] Figure 3 is the wave shape display of the wave gather recording of the double-layer single-interface model and its acoustic wave forward modeling in embodiment 1 after cutting off the direct wave;

[0081] Figure 4 is the wave gather recording of the double-layer single-interface model and its acoustic wave forward modeling in embodiment 1 after cutting off the direct wave;

[0082] Figure 5 is the wave gather recording of the double-layer single-interface model and its acoustic wave forward modeling in embodiment 1 after cutting off the direct wave;

[0083] Figure 6 is the wave gather recording of the double-layer single-interface model and its acoustic wave forward modeling in embodiment 1 after cutting off the direct wave;

[0084] Figure 7 is the result of the layer interface gather processing after the conventional NMO correction in Example 2;

[0085] Figure 8 is the result of the layer interface gather processing after the amplitude-preserved frequency-raising processing based on the complex field Wiener filter in Example 2;

[0086] Figure 9 is the comparison chart of the target trace and the 50th trace after the filtering in Example 1;

[0087] Figure 10 is the trace spectrum of the original shot gather reflection event in Example 1;

[0088] Figure 11 is the trace spectrum of the reflection event after the conventional NMO correction in Example 1;

[0089] Figure 12 is the trace spectrum of the reflection event after the amplitude-preserved frequency-raising processing based on the complex field Wiener filter in Example 1;

[0090] Figure 13 is the comparison chart of the 50th trace of the shot gather and the 50th trace after the amplitude compensation filtering in Example 1;

[0091] Figure 14 is the five-layer four-interface model and the acoustic forward shot record thereof in Example 1;

[0092] Figure 15 is the shot record after the direct wave is cut off in the five-layer four-interface model and the acoustic forward thereof in Example 1(a);

[0093] Figure 15 is the result of the layer interface gather processing after the conventional NMO correction in Example 1 and 2;

[0094] Figure 15 is the gather record after the conventional gather distortion cutting processing in Example 2(c);

[0095] Figure 15 is the gather record after the amplitude-preserved frequency-raising processing based on the complex field Wiener filter in Example 1 and 2(d). DETAILED DESCRIPTION

[0096] The drawings are only for illustrative purposes and should not be construed as limiting the patent; in order to better illustrate the embodiments, some components in the drawings may be omitted, enlarged or reduced, and do not represent the actual product size; it is understandable for those skilled in the art that some well-known structures and their descriptions in the drawings may be omitted. The positional relationship described in the drawings is only for illustrative purposes and should not be construed as limiting the patent.

[0097] In the accompanying drawings of the embodiments of the present invention, the same or similar reference numerals correspond to the same or similar components. In the description of the present invention, it should be understood that if terms such as "upper," "lower," "left," "right," "long," and "short" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the terms used to describe positional relationships in the drawings are only for illustrative purposes and should not be construed as limiting the present patent. For those skilled in the art, the specific meaning of the above terms can be understood according to the specific circumstances.

[0098] The technical solution of the present invention will be further described in detail below through specific embodiments and in conjunction with the accompanying drawings:

[0099] Example 1

[0100] In this embodiment, as Figure 1 As shown, a gather-preserving, frequency-enhancing, and detuning correction method based on complex-domain Wiener filtering includes the following:

[0101] S1: Construct the analytic signal in the complex domain of the pre-stack gather and obtain the maximum instantaneous amplitude of the pre-stack gather;

[0102] S2: Select the target trace from the pre-stack trace set in step S1;

[0103] S3: Based on the target channel in step S2, Wiener filtering is performed on the remaining channels in the pre-stack channel set except for the target channel by obtaining the Wiener filtering operator of the analytic signal in step S1.

[0104] S4: Calculate the maximum instantaneous amplitude of the pre-stack gather after Wiener filtering in step S3, and determine the ratio coefficient between the maximum instantaneous amplitude of the pre-stack gather before Wiener filtering and the maximum instantaneous amplitude after Wiener filtering.

[0105] S5: The required pre-stack gather is obtained by performing AVO amplitude compensation on the Wiener-filtered pre-stack gather in step S4.

[0106] In this embodiment, by constructing an analytical signal, selecting the target trace, and applying Wiener filtering to the remaining traces in the pre-stack trace set, amplitude compensation is then performed on the nanofiltered pre-stack trace set to obtain the desired pre-stack trace set. This embodiment avoids the drawbacks of traditional trace set distortion removal methods, such as the loss of shallow and long-range seismic data, loss of shallow geological information, and artificial reduction of shallow cover times. Meanwhile, as... Figure 2 and Figure 14As shown, this embodiment also constructs a two-layer single-interface model and a five-layer four-interface model. The two-layer single-interface model can effectively verify the processing effect of the stretching of the long-range gun-detector waveform. The five-layer four-interface model can effectively verify the waveform aliasing and stretching processing effect of the thin interlayer tuning effect.

[0107] The analytic signal for constructing the complex field of the trace collection in step S1 includes the following:

[0108] Assuming the total number of channels in the pre-stack gather is n, then the pre-stack gather is:

[0109] S(t)=[s1(t),s2(t),s3(t),…,s n (t)] (1);

[0110] Perform a Hilbert transform on any trace in the pre-stack trace set:

[0111]

[0112] The analytical signal expression for this channel is:

[0113]

[0114] The analytic signal in the complex field of the pre-stack trace set is:

[0115] Z(t)=[z1(t),z2(t),z3(t),…,z n (t)] (4);

[0116] In the formula, S(t) is a function of the pre-stack gather with respect to time, s n (t) is a function of time for the nth seismic trace. Let s(t) be a function of time after the Hilbert transform, z(t) be a function of time for each analytic signal, and Z(t) be a function of time for the analytic signal of the pre-stack gather. Here, t is time, τ is a mathematical variable, and i is the imaginary unit. In this embodiment, the analytic signal essentially transforms a narrowband real signal into a complex signal, and the Hilbert transform can generate a unique complex signal for a narrowband real signal. In this embodiment, as... Figure 3 As shown, in the process of converting shot gathers to pre-stack gathers, the direct waves are first removed. After obtaining the interface containing only the descriptive layer interface, we perform a Hilbert transform on each signal in S(t). Taking the vertical incident channel s(t) of a zero-shot-receiver offset seismic wave as an example, the result after aligning and performing the Hilbert transform is as follows. Figure 4 The dashed line shows the signal after the Hilbert transform. The phase difference with the original signal is 90°. For the gather S(t), the signal after each Hilbert transform is... As the imaginary part, the original signal s(t) as the real part, the complex domain analytic signal trace set is constructed as Figure 5 as shown (so the trace set is complex domain, Figure 5 only show the imaginary part of the analytic signal).

[0117] In addition, the maximum instantaneous amplitude of the pre-stack trace set in step S1 includes the following:

[0118] The instantaneous amplitude of any point in each trace analytic signal is:

[0119]

[0120] The instantaneous amplitude of the pre-stack trace set is:

[0121] A(t) = [a1(t), a2(t), a3(t), …, a n (t)] (6);

[0122] In the formula, a(t) is the instantaneous amplitude of any point in each trace analytic signal, s(t) is a function of time for each seismic trace, is a function of time after the Hilbert transform of s(t), A(t) is the instantaneous amplitude of the pre-stack trace set, a n (t) is the instantaneous amplitude of any point in the nth trace analytic signal.

[0123] Wherein, the step S3 of calculating the Wiener filter operator of the analytic signal includes the following:

[0124] Assuming that the analytic signal z1(t) of the first seismic trace in Z(t) = [z1(t), z2(t), z3(t), …, z n (t)] is the target trace, the analytic signal z2(t) of the second seismic trace is the to-be-matched processing trace, and the filtering expectation of the analytic signal z1(t) of the first seismic trace is constantly approaching the analytic signal z2(t) of the second seismic trace, then the actual expected output of z2(t) can be expressed as:

[0125]

[0126] The expected output

[0127] In the formula, h(t) is a filter operator, is expressed as the extension of the Wiener filter, and t is time. In this embodiment, the Wiener filter is the optimal filter under the minimum mean square criterion based on statistical theory. The Wiener filter is suitable for constructing a matched filter operator because it can be used as a linear invariant filter to seek an output that approximates the expected signal.

[0128] In addition, the expected output Also expressed in integral form:

[0129]

[0130] The expected error of the analytic signal z2(t) of the second seismic trace is:

[0131]

[0132] The total energy mean square error of the analytic signal z2(t) of the second seismic trace is:

[0133]

[0134] where h(t) is a filter operator, τ is a mathematical variable, z1(t) is the analytic signal of the first seismic trace, z2(t) is the analytic signal of the second seismic trace, expressed as an extension of the Wiener filter.

[0135] The total energy mean square error of the analytic signal z2(t) of the second seismic trace can also be expressed according to the equations (9) to (11) as:

[0136]

[0137] changing the integral variable:

[0138]

[0139] The total energy mean square error of the analytic signal z2(t) of the second seismic trace is calculated according to the equations (12) and (13) as:

[0140]

[0141] where ε(t) is the expected error of the analytic signal z2(t) of the second seismic trace, E[ε 2 (t)] is the total energy mean square error of the analytic signal z2(t) of the second seismic trace, expressed as an extension of the Wiener filter, h(t) is a filter operator, t is time, z1(t) is the analytic signal of the first seismic trace, z2(t) is the analytic signal of the second seismic trace, τ is a mathematical variable, v is a mathematical variable, expressed as the autocorrelation of z1(t), expressed as the cross-correlation of z1(t) and z2(t), expressed as the autocorrelation of z2(t).

[0142] In addition, the total energy mean square error of the analytic signal z2(t) of the second seismic trace is calculated by partial derivation and the partial derivation is set to zero: ​

[0143]

[0144] Meanwhile, the mathematical expression of the matched filter operator can be expressed as a linear equation set:

[0145]

[0146] The matrix form of equation (16) is:

[0147]

[0148] In the equation, is the operator matrix of each trace in the pre-stack gather;

[0149] The operator matrix of the whole pre-stack gather is:

[0150]

[0151] In the equation, ε(t) is the expected error of the analytic signal z2(t) of the second seismic trace, E[ε 2 (t)] is the total energy mean square error of the analytic signal z2(t) of the second seismic trace, is represented as the extension of the Wiener filter, h(t) is the filter operator, t is time, z1(t) is the analytic signal of the first seismic trace, z2(t) is the analytic signal of the second seismic trace, τ is a mathematical variable, and v is a mathematical variable, is the autocorrelation of z1(t), is the cross-correlation of z1(t) and z2(t), is the autocorrelation of z2(t), m represents the number of rows or columns of a matrix, and H represents the operator matrix of the whole pre-stack gather. In the embodiment, Figure 9 is a comparison chart of the target trace and the 50th filtered trace, Figure 9 The realization in the figure indicates the target trace, and the dashed line indicates the 50th filtered trace, from which it can be seen that the filtered result of the 50th seismic trace is infinitely close to the vertical incidence trace of the seismic wave at zero offset. Figure 9 Figure 10 is the spectrum of each trace of the original shot gather reflection event, Figure 11 is the spectrum of each trace of the conventional NMO corrected reflection event, Figure 12 is the spectrum of each trace of the reflection event after the amplitude-preserved frequency-raising processing based on the complex domain Wiener filter. It can be seen that the main frequency of each trace of the original shot gather remains basically unchanged, while the amplitude of the trace set after the conventional NMO processing keeps increasing, but the frequency keeps moving to the low frequency end, which leads to the stretching of the waveform of the far trace. When thin interbeds are encountered, the tuning effect of the trace set is prone to occur (such as Figure 15 ​b), and the method for preserving amplitude and increasing frequency of gathers based on complex domain Wiener filter in the present application, after the Wiener filter is performed by solving the Wiener filter operator, each gather of the processed gathers is basically consistent in near and far frequencies, compared with the conventional processing result, the frequency of the far gather is increased, thereby the waveform stretching phenomenon of large offset is eliminated, and the effect of de-tuning is achieved when thin interbeds are processed (for example Figure 15 d).

[0152] In the step S4, the maximum instantaneous amplitude of the pre-stack gather after the Wiener filter is solved, including the following:

[0153] The analytic signal of the pre-stack gather in the complex domain after the Wiener filter is:

[0154]

[0155] The instantaneous amplitude sequence of each analytic signal of the analytic signal of the pre-stack gather in the complex domain after the Wiener filter is calculated

[0156]

[0157]

[0158] The amplitude proportion coefficient sequence c(t) of each gather before and after the Wiener matching filter is calculated

[0159]

[0160] The proportion coefficient matrix of the pre-stack gather is:

[0161] C(t)=[1,c2(t),c3(t),…,c n (t)] (23);

[0162] In the formula, z1(t) represents the analytic signal of the first seismic gather, represents the actual expected output of the n-th seismic data, s(t) represents the function of each seismic gather with respect to time, is the function of s(t) after the Hilbert transform, t is time, i is the imaginary unit, z1(t) is the analytic signal of the first seismic gather, represents the function of any gather in the pre-stack gather after the Wiener filter with respect to time, is obtained by performing the Hilbert transform.

[0163] Embodiment 2

[0164] The present embodiment is similar to embodiment 1, and in the present embodiment, the AVO amplitude compensation of the pre-stack gather includes the following:

[0165] extract The real part is obtained as follows:

[0166]

[0167] Calculate the pre-stack gather S for amplitude preservation and frequency enhancement to eliminate tuning. Wiener (t):

[0168]

[0169] In the formula, z1(t) represents the analytical signal of the first seismic trace. Let s1(t) represent the expected output of the nth seismic trace, and let s1(t) represent the function of time for the first seismic trace. Let C(t) be the function of time for the nth seismic trace in the pre-stack gather after Wiener filtering, and let C(t) be the scaling factor matrix of the pre-stack gather. In this embodiment, Figure 6 For the phase axis of the reflected wave at the interface, Figure 7 The results of conventional NMO correction from shot gather to pre-stack gather show that, when dealing with single interfaces or very thick layer interfaces, conventional correction methods can achieve relatively good flattening under the premise of accurate velocity setting. However, at the far end of the shot receiver, the waveform of the in-phase axis of the reflected wave inevitably exhibits waveform stretching due to the decrease in frequency and time thickness. This is especially problematic when encountering thin interlayers ( Figure 15 When b), a tuning effect of the gather will occur. To address this problem, the traditional distortion removal method ( Figure 15 c) Although distorted data is eliminated, effective information about gun ranging is also reduced, which is considered to decrease the number of observations and limit the subsequent inversion and identification capabilities. The gather amplitude-preserving, frequency-enhancing, and detuning correction method based on Wiener filtering in the complex domain proposed in this invention, however, has better processing results ( Figure 8 , Figure 15 d) It effectively eliminates the tuning effect generated by the gather, and under the premise of amplitude preservation and frequency enhancement, it is suitable for single-interface or very thick-layer interfaces. Figure 8 The in-phase axis can achieve a more horizontal reflected wave in-phase axis without requiring a given velocity. For complex geological conditions and thin interlayered doped interfaces, it reduces waveform stretching distortion while preserving effective information, greatly improving seismic and visual resolution, and achieving effective detuning. In this embodiment, the compensated gather is as follows: Figure 13 As shown, where Figure 13 The dashed line represents the target trace, and the dashed line represents the trace gather after amplitude compensation (which is close to coinciding with the solid line).

[0170] Example 3

[0171] Similar to the embodiment 1, in the present embodiment, the target trace is the seismic trace with high frequency of offset in the prestack gather. In the present embodiment, the generation of the gather tuning effect is mainly due to the change of the time thickness and the seismic frequency. The large offset is prone to produce the waveform stretch distortion due to the decrease of the time thickness and the low frequency component. The recorded data of the target trace prestack gather at the near offset is obvious in the reflection layer, and the information of the target layer is well reflected. The generation of the gather tuning effect is mainly due to the change of the time thickness and the seismic frequency. The large offset is prone to produce the waveform stretch distortion due to the decrease of the time thickness and the low frequency component. In order to correct the weakening and even eliminate the influence, the target trace is preferably selected at the near offset.

[0172] Obviously, the above embodiments of the present application are only examples for clearly illustrating the present application, and are not intended to limit the embodiments of the present application. Based on the above description, other different forms of changes or variations can be made by those skilled in the art. Here, all the embodiments are not required to be exhausted. Any modification, equivalent replacement and improvement within the spirit and principle of the present application should be included in the protection scope of the claims of the present application.

Claims

1. A gather-preserving, frequency-enhancing, and detuning correction method based on complex-domain Wiener filtering, characterized in that, Including the following: S1: Construct the analytic signal in the complex domain of the pre-stack gather and obtain the maximum instantaneous amplitude of the pre-stack gather; S2: Select the target trace from the pre-stack trace set described in step S1; S3: Based on the target channel in step S2, Wiener filtering is performed on the remaining channels in the pre-stack channel set except for the target channel by obtaining the Wiener filtering operator of the analytical signal described in step S1. S4: Calculate the maximum instantaneous amplitude of the pre-stack gather after Wiener filtering in step S3, and determine the ratio coefficient between the maximum instantaneous amplitude of the pre-stack gather before Wiener filtering and the maximum instantaneous amplitude after Wiener filtering. S5: The required pre-stack gather is obtained by performing AVO amplitude compensation on the Wiener-filtered pre-stack gather in step S4; The analytic signal for constructing the complex domain of the pre-stack trace set mentioned in step S1 includes the following: Assuming the total number of channels in the pre-stack gather is n, then the pre-stack gather is: （1）； Perform a Hilbert transform on any trace in the pre-stack trace set: （2）； The analytical signal expression for this channel is: （3）； The analytic signal in the complex field of the pre-stack trace set is: （4）； In the formula, S(t) is a function of the pre-stack gather with respect to time. Let n be a function of time for the nth seismic trace. for The function of time after the Hilbert transform, The analytic signal for each seismic trace is a function of time. Let t be the analytic signal of the pre-stack gather as a function of time, where t is time, τ is a mathematical variable, and i is the imaginary unit.

2. The gather amplitude-preserving, frequency-enhancing, and detuning correction method based on complex-domain Wiener filtering according to claim 1, characterized in that, The step S1 of determining the maximum instantaneous amplitude of the pre-stack gather includes... the following: The instantaneous amplitude at any point in each analytical signal is: （5）； The instantaneous amplitude of the pre-stack set is: （6）； In the formula, The instantaneous amplitude at any point in each analytical signal. For each seismic trace, it is a function of time. for The function of time after the Hilbert transform, The instantaneous amplitude of the pre-stack set, Let be the instantaneous amplitude at any point in the nth analytic signal.

3. The gather amplitude-preserving, frequency-enhancing, and detuning correction method based on complex-domain Wiener filtering according to claim 2, characterized in that, The Wiener filter operator for obtaining the analytic signal in step S3 includes the following: Assumption Analytical signal of the first seismic trace within Analytical signal of the second seismic trace, for the target trace. For the traces to be matched, the analytical signal of the first seismic trace is used. Analytical signal for the second seismic trace As the filtering expectation continues to approach, Actual expected output It can be represented as: （7）； Expected output (8); In the formula, For filtering operators, This is represented as an extension of the Wiener filter, where t is time.

4. The gather amplitude-preserving, frequency-enhancing, and detuning correction method based on complex-domain Wiener filtering according to claim 3, characterized in that, The expected output Represented in integral form: （9）； The analytical signal of the second seismic trace The expected error is: （10）； Analytical signals from the second seismic trace The total energy mean square error is: （11）； In the formula, Let τ be a filtering operator, and τ be a mathematical variable. This is the analytical signal of the first seismic trace. This is the analytical signal of the second seismic trace. This is represented as an extension of Wiener filtering.

5. The gather amplitude-preserving, frequency-enhancing, and detuning correction method based on complex-domain Wiener filtering according to claim 4, characterized in that, According to equations (9) to (11), the analytical signal of the second seismic trace The total energy mean square error is expressed as: （12）； Change Integral variables: （13）； The analytical signal of the second seismic trace is calculated based on equations (12) and (13). The total energy mean square error is: （14）； In the formula, Analytical signal for the second seismic trace The expected error Analytical signal for the second seismic trace The total energy mean square error, This is represented as an extension of the Wiener filter. Here, t is the filtering operator, and t is time. This is the analytical signal of the first seismic trace. Let be the analytical signal of the second seismic trace, τ be a mathematical variable, and v be a mathematical variable. Represented as autocorrelation, express and The interrelation, Represented as Autocorrelation.

6. The gather amplitude-preserving, frequency-enhancing, and detuning correction method based on complex-domain Wiener filtering according to claim 5, characterized in that, Analytical signals of the second seismic trace The partial derivative of the total energy mean square error is calculated and set to zero: (15)； Meanwhile, the mathematical expression for the matched filter operator is represented by a system of linear equations: （16）； The matrix form of equation (16) is: (17)； In the formula, To match the operator matrix of the second track in the pre-stack track set; The operator matrix for the entire pre-stack gather is then: （18）； In the formula, Analytical signal for the second seismic trace The expected error Analytical signal for the second seismic trace The total energy mean square error, This is represented as an extension of the Wiener filter. Here, t is the filtering operator, and t is time. This is the analytical signal of the first seismic trace. Let be the analytical signal of the second seismic trace, τ be a mathematical variable, and v be a mathematical variable. Represented as autocorrelation, express and The interrelation, Represented as The autocorrelation, where m represents the number of rows or columns of the matrix. It is represented as the operator matrix of the entire pre-stack set.

7. The gather amplitude-preserving, frequency-enhancing, and detuning correction method based on complex-domain Wiener filtering according to claim 6, characterized in that, The step S4 for determining the maximum instantaneous amplitude of the pre-stack gather after Wiener filtering includes the following: The analytic signal in the complex domain of the pre-stack gather after Wiener filtering is: （19）； Calculate the instantaneous amplitude sequence of each analytic signal in the complex domain of the pre-stack gather after Wiener filtering. : （20）； （21）； Calculate the amplitude scaling factor sequence for each channel before and after Wiener matched filtering. : （22）； The scaling factor matrix of the pre-stack gather is: （23）； In the formula, This is represented as the analytical signal of the first seismic trace. This represents the actual expected output of the nth seismic data. This is expressed as a function of time for each seismic trace. for The function of time after the Hilbert transform, where t is time and i is the imaginary unit. This is the analytical signal of the first seismic trace. It can be expressed as a function of time for any seismic trace in the pre-stack trace set after Wiener filtering. for The result obtained by performing a Hilbert transform.

8. The gather amplitude-preserving, frequency-enhancing, and detuning correction method based on complex-domain Wiener filtering according to claim 7, characterized in that, The pre-stack gather AVO amplitude compensation in step S5 includes the following: extract The real part is obtained as follows: （24）； Calculate the pre-stack gather for amplitude preservation and frequency boosting to eliminate tuning. : （25）； In the formula, This is represented as the analytical signal of the first seismic trace. This represents the actual expected output of the nth seismic data. It is expressed as a function of time for the first seismic trace. It is expressed as a function of time for the nth seismic trace in the pre-stack trace set after Wiener filtering. It is represented as the scaling factor matrix of the pre-stack gather.

9. The gather-preserving amplitude-raising frequency-enhancing detuning correction method based on complex-domain Wiener filtering according to claim 1, characterized in that... The target trace in step S2 is a seismic trace with a high shot-receiver offset frequency within the pre-stack trace set.

Images