Seismic velocity signal and acceleration signal combined high-resolution processing method

Abstract

The invention provides a seismic velocity signal and acceleration signal combined high-resolution processing method. The seismic velocity signal and acceleration signal combined high-resolution processing method comprises the following steps: step 1, carrying out combined processing on a seismic velocity signal and an acceleration signal; 2, signal decomposition based on independent component analysis is carried out; step 3, seismic signal joint high-resolution processing based on Kalman optimal estimation is carried out; and 4, inputting various signals to obtain a combined high-resolution processing result. According to the seismic velocity signal and acceleration signal combined high-resolution processing method, the problem that dependence on selected signals is large due to the fact that only one signal is used for high-resolution processing at present is solved, and the resolution and accuracy of seismic signals are improved through combined processing of velocity signals and acceleration signals.

Description

Technical Field

[0001] The invention relates to the technical field of petroleum geophysical exploration, and in particular to a method for combining seismic velocity signals and acceleration signals with high resolution. Background Art

[0002] In recent years, my country's oil and gas exploration is in an important transition period. The change of exploration targets and exploration technologies has gradually increased the difficulty of exploration. The exploration targets have gradually changed from large, thick and strong oil and gas reservoirs to small, thin and weak oil and gas reservoirs, and from structural oil and gas reservoirs to complex oil and gas reservoirs such as stratigraphy and lithology. In addition, oil and gas exploration has also mainly changed to thin-layer detection. Therefore, we need to explore processing technologies with higher resolution and better accuracy.

[0003] Compared with velocity signals, acceleration signals have great advantages in terms of bandwidth and octave; velocity signals have more effective information in the low-frequency part compared with acceleration signals. Therefore, high-resolution processing of the combined velocity signal and acceleration signal can not only retain factors such as bandwidth, but also retain the low-frequency effective signal. In short, the combined high-resolution processing method of seismic velocity signals and acceleration signals is a feasible and necessary method.

[0004] In the Chinese patent application with application number: CN201710800277.6, a method, device and system for determining stacking velocity are involved, the method includes obtaining seismic reflection data of seismic records, calculating the slope of the local event axis according to the seismic reflection data; calculating the relationship data between zero offset time and stacking velocity according to the slope of the local event axis; calculating the local similarity coefficient between the determined reference track and the seismic track corresponding to the relationship data; determining the similarity coefficient of the relationship data according to the local similarity coefficient between the reference track and the seismic track corresponding to the relationship data; using the similarity coefficient of the relationship data as the weight of the relationship data, calculating the center point of the weighted relationship data point according to the weighted clustering algorithm; determining the stacking velocity of the seismic record according to the velocity value of the center point. Using the various embodiments of this application, the accuracy of the obtained stacking velocity can be improved.

[0005] In the Chinese patent application with application number: CN201410364709.X, an adaptive high-precision covariance regularized seismic data stacking velocity analysis method is involved, including the following steps: A. inputting all CMP gathers after preprocessing; B. setting the scanning speed range and scanning spacing, and selecting a scanning speed to perform normal time difference correction scanning on all CMP gathers in sequence and obtain all corrected CMP gathers; C. selecting a corrected CMP gather from all the corrected CMP gathers, and dividing the one corrected CMP gather into multiple time windows, obtaining the covariance matrix of the gathers in one of the multiple time windows, regularizing the covariance matrix of the gathers in the one time window and obtaining the values ​​of the velocity spectrum of all the sample points in the one time window by setting a threshold value of the signal-to-noise ratio coefficient; D. sliding the time window until the multiple time windows are processed and the values ​​of the velocity spectrum of all the sample points in the one corrected CMP gather are obtained.

[0006] In the Chinese patent application with application number: CN201210017116.7, a method for optimizing seismic velocity analysis data for oil and gas exploration is involved, which includes performing velocity gather sorting on seismic data to sort out super gathers for velocity analysis; performing frequency band abnormal amplitude attenuation; performing FXY domain three-dimensional pre-stack random noise attenuation; performing frequency filtering to enhance the energy of the dominant frequency band of the reflected wave; and performing wavelet shaping deconvolution processing to improve the resolution. The method for optimizing seismic velocity analysis data for oil and gas exploration solves the problem of low focusing accuracy of velocity analysis energy clusters in the prior art, and has the advantages of enhancing the energy of effective reflection, improving the quality of velocity analysis data, improving the signal-to-noise ratio and resolution of velocity analysis gathers, and improving the focusing of velocity spectrum.

[0007] In the Chinese patent application with application number: CN201810486300.3, a new method for applying onshore seismic acceleration signals is involved, which includes: step 1, obtaining acceleration forward simulation seismic records, estimating the number of acceleration signal acquisition coverage, and acquiring seismic acceleration signals; step 2, suppressing and eliminating high-frequency noise; step 3, performing acceleration seismic signal imaging, and applying seismic data processing technology to the data after high-frequency noise suppression to complete acceleration seismic data imaging processing. The new method for applying onshore seismic acceleration signals effectively improves the imaging accuracy of seismic exploration, which is mainly manifested in the high fidelity, high resolution, wide bandwidth, and high sensitivity to changes in medium physical properties of seismic signals. It is an effective method to improve the accuracy of seismic exploration onshore.

[0008] The above existing technologies are greatly different from the present invention and fail to solve the technical problem we want to solve. For this reason, we have invented a new high-resolution processing method for combining seismic velocity signals and acceleration signals. Summary of the invention

[0009] The purpose of the present invention is to provide a method for joint high-resolution processing of seismic velocity signals and acceleration signals, which solves the problem of currently using only one seismic signal for high-resolution processing and realizes joint processing of seismic velocity signals and acceleration signals.

[0010] The object of the present invention can be achieved by the following technical measures: a method for combining high-resolution processing of seismic velocity signals and acceleration signals, the method for combining high-resolution processing of seismic velocity signals and acceleration signals comprising:

[0011] Step 1, performing joint processing of seismic velocity signals and acceleration signals;

[0012] Step 2, perform signal decomposition based on independent component analysis;

[0013] Step 3, performing joint high-resolution processing of seismic signals based on Kalman optimal estimation;

[0014] Step 4: Input multiple signals to obtain the result of joint high-resolution processing.

[0015] The purpose of the present invention can also be achieved by the following technical measures:

[0016] Step 1 includes:

[0017] Step 11, converting the seismic velocity signal into an acceleration signal, and normalizing it according to the acceleration signal;

[0018] Step 12, standardizing the acceleration signal converted from the velocity signal and the acceleration signal;

[0019] Step 13, whiten the standardized data.

[0020] In step 11, the specific calculation formula is as follows:

[0021] X vel-acc =X′ vel X acc_max / X vel-acc_max

[0022] Among them, X vel-acc is the acceleration signal after the velocity signal is converted and normalized, X acc_max is the maximum value of the acceleration signal, X vel-acc_max is the maximum value of the converted acceleration signal, X vel is the speed signal, and the ' in the upper right corner represents the derivative operation of the signal.

[0023] In step 12, the normalization formula is as follows:

[0024] Xi，j =X i，j -Xm j

[0025] Where X is the seismic record, i.e. the acceleration signal, or the acceleration signal converted from the velocity signal, i is the number of rows, i.e. the sampling points, j is the number of columns, i.e. the number of channels, Xm j is the mean of the j-th data.

[0026] In step 13, the whitening formula is as follows:

[0027] Xw=D -1 I -1 X

[0028] Among them, Xw is the whitened data, D is the eigenvalue of the covariance data of the standardized data S, I is the unit matrix, S is the standardized data, and -1 means to invert the matrix.

[0029] Step 2 includes:

[0030] Step 21, given the initial value matrix B, solve the mixing matrix;

[0031] Step 22, determine whether the solved matrix B converges;

[0032] Step 23, reconstructing the data using the unmixing matrix, can obtain mutually independent separated data, namely, effective signal and noise data.

[0033] In step 21, the formula for solving the mixing matrix is:

[0034] S=B -1 X w

[0035] Among them, S is the source signal, and B is the estimation obtained after analyzing the source signal.

[0036] In step 22, if the solved matrix B converges, the unmixing matrix is ​​obtained; if not, the process returns to step 1 until it converges.

[0037] Step 3 includes:

[0038] Step 31, calculating the variance Q of the noise obtained by independent component analysis;

[0039] Step 32, calculating the initial state estimate and the error covariance matrix;

[0040] Step 33, calculating the gain matrix Z;

[0041] Step 34: Update the state prediction and error covariance matrix according to the gain matrix Z.

[0042] In step 31, according to the characteristic that the seismic signal can be regarded as a convolution model, given the state matrix F, the observation matrix H, and the observation noise variance C, the variance Q of the noise obtained by independent component analysis is calculated.

[0043] In step 32, the calculation formula is as follows:

[0044] R(k+1)=FR(k)

[0045] P(k+1)=FP(k)F T +Q

[0046] Among them, R is the state estimation, which is composed of the acceleration signal converted from the velocity signal and the effective signal obtained by independent component analysis, P is the initial error covariance estimate, k is the number of iterations, and T in the upper right corner represents the inverse operation of the matrix.

[0047] In step 33, the formula for calculating the gain matrix Z is as follows:

[0048] Z(k)=P(k)H T (k)[H(k)P(k)H T (k)+C] -1 .

[0049] In step 34, the update formula is as follows:

[0050] R n (k+1)=R(k)+Z(k)[X(k)-HR(k)]

[0051] P n (k+1)=[IZ(k)H]P(k)

[0052] Among them, R n is the updated state estimate, P n is the updated error covariance matrix, and I is the identity matrix.

[0053] In step 4, if the updated state estimate and error covariance matrix meet the accuracy requirements, the calculation is stopped and the state estimate is output as the result of the Kalman optimal estimate, that is, the result after high-resolution processing; if the accuracy requirements are not met, step 3 is repeated until the accuracy requirements are met.

[0054] The method for combined high-resolution processing of seismic velocity signals and acceleration signals of the present invention belongs to the field of oil exploration, and includes: realizing the conversion of velocity signals into acceleration signals; realizing independent component analysis (ICA) of velocity signals and acceleration signals; realizing Kalman optimal estimation (Kalman) of velocity signals and acceleration signals; and obtaining high-resolution results by inputting post-stack data of velocity signals and acceleration signals. In view of some shortcomings existing in conventional high-resolution methods, the present invention proposes a method for combined processing of velocity signals and acceleration signals. It mainly includes: realizing the decomposition of seismic signals through independent component analysis technology, so that independent components are separated; using Kalman optimal estimation to process the results of independent component analysis processing and the combined signal, realizing high-resolution processing of seismic acceleration signals, that is, improving the time resolution and the bandwidth of acceleration signals.

[0055] The invention solves the problem of currently using only one type of seismic signal for high-resolution processing, realizes the joint processing of seismic velocity signals and acceleration signals, improves the resolution of complex geological bodies, and is applicable to the joint high-resolution processing of multiple signals. The beneficial effects of the invention are: the invention provides a method for joint high-resolution processing of seismic velocity signals and acceleration signals, which is used to solve the problem of currently using only one type of signal for high-resolution processing, and thus has a high dependence on the selected signal, and utilizes the joint processing of velocity signals and acceleration signals to improve the resolution and accuracy of seismic signals. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 A flowchart of a specific embodiment of the method for combining high-resolution processing of seismic velocity signals and acceleration signals of the present invention;

[0057] Figure 2 Schematic diagram of seismic signals in a specific embodiment of the present invention: (a) velocity signal, (b) acceleration signal;

[0058] Figure 3 A schematic diagram of an acceleration signal converted from a seismic velocity signal in a specific embodiment of the present invention;

[0059] Figure 4 Schematic diagram of mutually independent results obtained by performing independent component analysis using acceleration signals alone in a specific embodiment of the present invention: (a) effective signal, (b) noise signal;

[0060] Figure 5 Schematic diagram of mutually independent results obtained by performing independent component analysis using a velocity signal and an acceleration signal in combination in a specific embodiment of the present invention: (a) effective signal, (b) noise signal;

[0061] Figure 6 A schematic diagram of initial parameters of Kalman optimal estimation in a specific embodiment of the present invention;

[0062] Figure 7 A schematic diagram of a high-resolution result obtained by performing Kalman optimal estimation using acceleration signals alone in a specific embodiment of the present invention;

[0063] Figure 8 It is a schematic diagram of a high-resolution result obtained by performing Kalman optimal estimation by jointly using a velocity signal and an acceleration signal in a specific embodiment of the present invention;

[0064] Fig. 9 Schematic diagram of time-frequency information of seismic acceleration signal in a specific embodiment of the present invention: (a) time-frequency of original acceleration signal, (b) time-frequency of high-resolution result obtained by separate processing, (c) time-frequency of high-resolution result obtained by joint processing;

[0065] Fig.10 The local images of the seismic acceleration signal in a specific embodiment of the present invention are: (a) a local original acceleration signal, (b) a local high-resolution result obtained by separate processing, and (c) a local high-resolution result obtained by combined high-resolution processing.

[0066] Fig.11 Schematic diagram of seismic signals in a specific embodiment 2 of the present invention: (a) velocity signal, (b) acceleration signal;

[0067] Fig.12 It is a schematic diagram of the high-resolution result in a specific embodiment 2 of the present invention;

[0068] Fig.13 It is a schematic diagram of initial parameters of Kalman optimal estimation in a specific embodiment 2 of the present invention;

[0069] Fig.14 Schematic diagram of time-frequency information of earthquake acceleration signal in a specific embodiment 2 of the present invention: (a) time-frequency of original acceleration signal, (b) time-frequency of high-resolution result;

[0070] Fig.15 Schematic diagram of seismic signals in a third specific embodiment of the present invention: (a) velocity signal, (b) acceleration signal;

[0071] Fig.16 It is a schematic diagram of the high-resolution result in a specific embodiment 3 of the present invention;

[0072] Fig.17 It is a schematic diagram of initial parameters of Kalman optimal estimation in a specific embodiment 3 of the present invention;

[0073] Fig.18 Schematic diagram of the time-frequency information of the earthquake acceleration signal in a specific embodiment 3 of the present invention: (a) the time-frequency of the original acceleration signal, and (b) the time-frequency of the high-resolution result. DETAILED DESCRIPTION

[0074] 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.

[0075] 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.

[0076] Conventional high-resolution processing methods often use only one type of data and are more dependent on the signal-to-noise ratio of the selected data. Compared with velocity signals, acceleration signals have great advantages in terms of bandwidth and octave; velocity signals have more effective information in the low-frequency part compared with acceleration signals. Therefore, high-resolution processing of the combined velocity signal and acceleration signal can not only retain factors such as bandwidth, but also retain low-frequency effective signals. In view of some shortcomings in the use of separate signals for high-resolution processing, as well as the advantages of seismic velocity signals and acceleration signals, the present invention proposes a method for combining seismic velocity signals and acceleration signals for high-resolution processing. It mainly includes: using independent component analysis (ICA) to obtain independent effective signals and noise signals of seismic signals; using Kalman optimal estimation to perform high-resolution processing on the obtained effective signals and original signals, thereby achieving high-resolution processing of seismic signals. Therefore, the present invention adopts high-resolution processing of seismic velocity signals and acceleration signals, avoiding the drawbacks of heavy reliance on selected data, and solving such problems well.

[0077] The technical solution of the present invention provides a method for joint high-resolution processing of seismic velocity signals and acceleration signals, including: realizing signal decomposition based on independent component analysis; realizing high-resolution processing based on Kalman optimal estimation; inputting multiple signals to obtain the result of joint high-resolution processing.

[0078] The present invention implements a signal decomposition algorithm based on independent variable analysis. First, the seismic velocity signal is converted into an acceleration signal and normalized according to the acceleration signal. The specific calculation formula is as follows:

[0079] X vel-acc =X′ vel X acc_max / X vel-acc_ m ax

[0080] Among them, X vel-acc is the acceleration signal after the velocity signal is converted and normalized, X acc_max is the maximum value of the acceleration signal, X vel-acc_max is the maximum value of the converted acceleration signal, X vel It is the speed signal, the upper right corner, represents the derivative operation of the signal.

[0081] Secondly, the acceleration signal converted from the velocity signal and the acceleration signal are standardized. The standardization formula is as follows:

[0082] X i，j =X i，j -Xm j

[0083] Where X is the seismic record (acceleration signal, or acceleration signal converted from velocity signal), i is the number of rows (sampling points), j is the number of columns (channels), and Xm j is the mean of the j-th data.

[0084] Again, the standardized data is whitened. The whitening formula is as follows:

[0085] Xw=D -1 I -1 X

[0086] Where Xw is the whitened data, D is the eigenvalue of the covariance data of the standardized data S, I is the unit matrix, S is the standardized data, and -1 indicates that the matrix is ​​inverted;

[0087] Next, given the initial value matrix B, solve the mixing matrix:

[0088] S=B -1 X w

[0089] Among them, S is the source signal, and B is the estimation obtained after analyzing the source signal.

[0090] Then, if the solved matrix B converges, the unmixing matrix is ​​obtained; if not, the above operation is repeated until it converges.

[0091] Finally, by reconstructing the data using the unmixing matrix, we can obtain independent separated data, namely, effective signal and noise data.

[0092] The combined high-resolution processing of seismic signals based on Kalman optimal estimation is realized. The effective signal obtained by independent component analysis is often inconsistent with the real seismic signal in terms of amplitude, phase, etc. By using Kalman optimal estimation, taking the obtained effective signal as the initial value, and performing high-resolution processing on the acceleration signal converted from velocity and the acceleration signal, these drawbacks can be avoided, and this type of problem is well solved. The present invention realizes high-resolution processing of seismic signals based on Kalman optimal estimation. First, according to the characteristic that the seismic signal can be regarded as a convolution model, the state matrix F, the observation matrix H, and the observation noise variance C are given, and the variance Q of the noise obtained by independent component analysis is calculated.

[0093] Secondly, calculate the initial state estimate and error covariance matrix, the calculation formula is as follows:

[0094] R(k+1)=FR(k)

[0095] P(k+1)=FP(k)FT+Q

[0096] Among them, R is the state estimation, which is composed of the acceleration signal converted from the velocity signal and the effective signal obtained by independent component analysis, P is the initial error covariance estimate, k is the number of iterations, and T in the upper right corner represents the inverse operation of the matrix.

[0097] Again, calculate the gain matrix Z, the calculation formula is as follows:

[0098] Z(k)=P(k)H T (k)[H(k)P(k)H T (k)+C] -1

[0099] Then, according to the gain matrix Z, the state estimation and error covariance matrix are updated, and the update formula is as follows:

[0100] R n (k+1)=R(k)+Z(k)[X(k)-HR(k)]

[0101] P n (k+1)=[IZ(k)H]P(k)

[0102] Among them, R n is the updated state estimate, p n is the updated error covariance matrix, and I is the identity matrix.

[0103] Finally, if the updated state estimate and error covariance matrix meet the accuracy requirements, the calculation is stopped and the state estimate is output as the result of the Kalman optimal estimate, that is, the result after high-resolution processing; if the accuracy requirements are not met, the above operation is repeated until the accuracy requirements are met.

[0104] The following are several specific embodiments of the present invention.

[0105] Example 1

[0106] In a specific embodiment 1 of the present invention, Figure 1 As shown, Figure 1 The flowchart of the method for processing seismic velocity signals and acceleration signals with high resolution of the present invention is as follows:

[0107] Joint processing of seismic velocity and acceleration signals;

[0108] Signal decomposition based on Independent Component Analysis (ICA);

[0109] Joint high-resolution processing of seismic signals based on Kalman optimal estimation.

[0110] The present invention realizes a joint high-resolution processing algorithm for velocity signals and acceleration signals. First, the seismic velocity signal is converted into an acceleration signal and normalized according to the acceleration signal. Secondly, the acceleration signal converted from the velocity signal and the acceleration signal are standardized. Thirdly, the standardized data is whitened. Next, the initial value matrix B is given and the mixing matrix is ​​solved. Then, if the solved matrix B converges, the unmixing matrix is ​​obtained. If not, the above operation is repeated until convergence. Next, the data is reconstructed using the unmixing matrix to obtain mutually independent separated data, that is, effective signals and noise data. Then, according to the characteristics that the seismic signal can be regarded as a convolution model, the state matrix F, the observation matrix H, and the observation noise variance C are given, and the variance Q of the noise obtained by independent component analysis is calculated. Secondly, the initial state estimate and the error covariance matrix are calculated. Thirdly, the gain matrix Z is calculated. Then, according to the gain matrix Z, the state estimate and the error covariance matrix are updated. Finally, if the updated state estimate and error covariance matrix meet the accuracy requirements, the calculation is stopped and the state estimate is output as the result of the Kalman optimal estimate, that is, the result after high-resolution processing; if the accuracy requirements are not met, the above operation is repeated until the accuracy requirements are met.

[0111] Example 2

[0112] In a specific embodiment 2 of the present invention, in order to verify the correctness of the present invention, a seismic signal of a certain work area (such as Figure 2 As shown in the figure), the time range is 0-2s, and the channel number range is 1-853. First, the velocity signal is converted into an acceleration signal and normalized (as shown in the figure). Figure 3As shown). If a conventional data is used for high-resolution processing, it is necessary to select two adjacent traces for separation when performing independent variable separation. Theoretically, it can be considered that the stratum reflection coefficients, that is, the geological structures, of two adjacent traces of data on the post-stack profile are similar and not very different. However, in actual data, there will be inconsistencies in adjacent traces in the seismic signal with a pinch-out point. At the same time, if one type of data is used for high-resolution processing, the advantages of multiple signals cannot be taken into account, which will greatly reduce the effect of high resolution. The best processing method is to use multiple signals for joint high-resolution processing, taking into account the advantages of multiple signals. For comparative analysis, the present invention uses two methods of high-resolution processing, namely, using acceleration signals alone and combining seismic velocity signals and acceleration signals, that is, using two methods to perform independent component analysis (such as Figure 4 ,like Figure 5 ) and Kalman optimal estimation (such as Figure 7 ,like Figure 8 ) to obtain the high-resolution result, the parameters in the Kalman optimal estimation are as follows Figure 6 The parameters shown are set, F is the state transfer matrix in the Kalman optimal estimation, H is the observation matrix in the Kalman optimal estimation, and C is the variance of the observation noise in the Kalman optimal estimation. Figure 7 and Figure 8 It can be seen that the structures obtained by the two methods are basically the same. Furthermore, the present invention compares the original acceleration signal and the high-resolution result by cutting out the range of 0-1s and 300-400 channels, as shown in FIG. Fig.10 As shown, it can be seen that the high-resolution results obtained by the two methods are basically consistent.

[0113] In order to verify the effectiveness of the present invention, the above raw acceleration signal and high-resolution results confirm that the results of the two methods are basically consistent, but there are great differences in details and resolution. The high-resolution results obtained by joint processing have better event axis continuity. Fig. 9 Shown are the time-frequency information of the original acceleration signal, the result of high-resolution using the acceleration signal alone, and the result of combined high-resolution processing. It can be seen that the longitudinal resolution (time resolution) of the combined high-resolution processing is higher.

[0114] Therefore, considering the disadvantages of using one type of data for high-resolution processing and the characteristic that high-resolution processing of multiple data can retain the advantages of different signals, the present invention proposes a method for high-resolution processing of seismic velocity signals and acceleration signals, providing a basis for high-resolution processing of multiple signals.

[0115] Example 3

[0116] In a specific embodiment 3 of the present invention, in order to verify the correctness of the present invention, a seismic signal of a certain work area (such as Fig.11As shown), the time range is 0-4s, and the number of channels range is 1-1762 channels. First, the velocity signal is converted into an acceleration signal and normalized. If a conventional data is used for high-resolution processing, it is necessary to select two adjacent channels for separation when performing independent variable separation. Theoretically, it can be considered that the stratum reflection coefficients, that is, the geological structures, of two adjacent channels of data on the post-stack profile are similar and not very different. However, in actual data, there will be inconsistencies in adjacent channels in seismic signals with vanishing points. At the same time, if one type of data is used for high-resolution processing, the advantages of multiple signals cannot be taken into account, which will greatly reduce the effect of high resolution. The best processing method is to use multiple signals for joint high-resolution processing, taking into account the advantages of multiple signals. The present invention adopts a method of combining seismic velocity signals and acceleration signals for high-resolution processing, that is, using two different signals to perform independent component analysis and Kalman optimal estimation to obtain high-resolution results (such as Fig.12 ), the parameters in the Kalman optimal estimation are as follows Fig.13 The parameters shown are set, F is the state transfer matrix in the Kalman optimal estimation, H is the observation matrix in the Kalman optimal estimation, and C is the variance of the observation noise in the Kalman optimal estimation. Fig.11 and Fig.12 It can be seen that the structures obtained before and after processing are basically the same, but the resulting event axes after processing are clearer and more continuous.

[0117] In order to verify the effectiveness of the present invention, the above raw acceleration signal and high-resolution results confirm that the results of this method are basically consistent. The high-resolution results obtained by joint processing have better continuity of the event axis. Fig.14 The figure shows the time-frequency information of the original acceleration signal and the result of the combined high-resolution processing. It can be seen that the longitudinal resolution (time resolution) of the data processed by this method is higher.

[0118] Example 4

[0119] In a specific embodiment 4 of the present invention, in order to verify the correctness of the present invention, a seismic signal of a certain work area (such as Fig.15As shown), the time range is 0-4s, and the number of channels range is 1-1768. First, the velocity signal is converted into an acceleration signal and normalized. If a conventional data is used for high-resolution processing, it is necessary to select two adjacent channels for separation when performing independent variable separation. Theoretically, it can be considered that the stratum reflection coefficients, that is, the geological structures, of two adjacent channels of data on the post-stack profile are similar and not very different. However, in actual data, there will be inconsistencies in adjacent channels in seismic signals with vanishing points. At the same time, if one type of data is used for high-resolution processing, the advantages of multiple signals cannot be taken into account, which will greatly reduce the effect of high resolution. The best processing method is to use multiple signals for joint high-resolution processing, taking into account the advantages of multiple signals. The present invention adopts a method of combining seismic velocity signals and acceleration signals for high-resolution processing, that is, using two different signals to perform independent component analysis and Kalman optimal estimation to obtain high-resolution results (such as Fig.16 ), the parameters in the Kalman optimal estimate are as follows Fig.17 The parameters shown are set, F is the state transfer matrix in the Kalman optimal estimation, H is the observation matrix in the Kalman optimal estimation, and C is the variance of the observation noise in the Kalman optimal estimation. Fig.15 and Fig.16 It can be seen that the structures obtained before and after processing are basically the same, but the resulting event axes after processing are clearer and more continuous.

[0120] In order to verify the effectiveness of the present invention, the above raw acceleration signal and high-resolution results confirm that the results of this method are basically consistent. The high-resolution results obtained by joint processing have better continuity of the event axis. Fig.18 The figure shows the time-frequency information of the original acceleration signal and the result of the combined high-resolution processing. It can be seen that the longitudinal resolution (time resolution) of the data processed by this method is higher.

[0121] 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.

[0122] 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 combining high-resolution processing of seismic velocity signals and acceleration signals, characterized in that: The method for combining high-resolution processing of seismic velocity signals and acceleration signals comprises: Step 1, performing joint processing of seismic velocity signals and acceleration signals; Step 2, perform signal decomposition based on independent component analysis; Step 3, performing joint high-resolution processing of seismic signals based on Kalman optimal estimation; Step 4: Input multiple signals to obtain the result of joint high-resolution processing.

2. The method for combined high-resolution processing of seismic velocity signals and acceleration signals according to claim 1, characterized in that: Step 1 includes: Step 11, converting the seismic velocity signal into an acceleration signal, and normalizing it according to the acceleration signal; Step 12, standardizing the acceleration signal converted from the velocity signal and the acceleration signal; Step 13, whiten the standardized data.

3. The method for combined high-resolution processing of seismic velocity signals and acceleration signals according to claim 2, characterized in that: In step 11, the specific calculation formula is as follows: X vel-acc =X′ vel X acc_max / X vel-acc_max Among them, X vel-acc is the acceleration signal after the velocity signal is converted and normalized, X acc_max is the maximum value of the acceleration signal, X vel-acc_max is the maximum value of the converted acceleration signal, X vel is the speed signal, and the ' in the upper right corner represents the derivative operation of the signal.

4. The method for combined high-resolution processing of seismic velocity signals and acceleration signals according to claim 3, characterized in that: In step 12, the normalization formula is as follows: X i，j =X i，j -Xm j Where X is the seismic record, i.e. the acceleration signal, or the acceleration signal converted from the velocity signal, i is the number of rows, i.e. the sampling points, j is the number of columns, i.e. the number of channels, Xm j is the mean of the j-th data.

5. The method for combining high-resolution processing of seismic velocity signals and acceleration signals according to claim 4, characterized in that: In step 13, the whitening formula is as follows: Xw=D -1 I -1 X Among them, Xw is the whitened data, D is the eigenvalue of the covariance data of the standardized data S, I is the unit matrix, S is the standardized data, and -1 means to invert the matrix.

6. The method for combined high-resolution processing of seismic velocity signals and acceleration signals according to claim 1, characterized in that: Step 2 includes: Step 21, given the initial value matrix B, solve the mixing matrix; Step 22, determine whether the solved matrix B converges; Step 23, reconstructing the data using the unmixing matrix, can obtain mutually independent separated data, namely, effective signal and noise data.

7. The method for combined high-resolution processing of seismic velocity signals and acceleration signals according to claim 6, characterized in that: In step 21, the formula for solving the mixing matrix is: S=B-1X Among them, S is the source signal, and B is the estimation obtained after analyzing the source signal.

8. The method for combined high-resolution processing of seismic velocity signals and acceleration signals according to claim 6, characterized in that: In step 22, if the solved matrix B converges, the unmixing matrix is ​​obtained; if not, the process returns to step 1 until it converges.

9. The method for combined high-resolution processing of seismic velocity signals and acceleration signals according to claim 1, characterized in that: Step 3 includes: Step 31, calculating the variance Q of the noise obtained by independent component analysis; Step 32, calculating the initial state estimate and the error covariance matrix; Step 33, calculating the gain matrix Z; Step 34: Update the state prediction and error covariance matrix according to the gain matrix Z.

10. The method for combined high-resolution processing of seismic velocity signals and acceleration signals according to claim 9, characterized in that: In step 31, according to the characteristic that the seismic signal can be regarded as a convolution model, given the state matrix F, the observation matrix H, and the observation noise variance C, the variance Q of the noise obtained by independent component analysis is calculated.

11. The method for combined high-resolution processing of seismic velocity signals and acceleration signals according to claim 10, characterized in that: In step 32, the calculation formula is as follows: R(k+1)FR(k) P(k+1)=FP(k)FT+Q Among them, R is the state estimation, which is composed of the acceleration signal converted from the velocity signal and the effective signal obtained by independent component analysis, P is the initial error covariance estimate, k is the number of iterations, and T in the upper right corner represents the inverse operation of the matrix.

12. The method for combined high-resolution processing of seismic velocity signals and acceleration signals according to claim 11, characterized in that: In step 33, the formula for calculating the gain matrix Z is as follows: Z(k)=P(k)H T (k)[H(k)P(k)H T (k)+C] -1 。 13. The method for combined high-resolution processing of seismic velocity signals and acceleration signals according to claim 12, characterized in that: In step 34, the update formula is as follows: R n (K+1)=R(K)+Z(K)[X(K)-HR(K)] P n (K+1)=[I-Z(K)H]P(K) Among them, R n is the updated state estimate, P n is the updated error covariance matrix, and I is the identity matrix.

14. The method for combined high-resolution processing of seismic velocity signals and acceleration signals according to claim 1, characterized in that: In step 4, if the updated state estimate and error covariance matrix meet the accuracy requirements, the calculation is stopped and the state estimate is output as the result of Kalman optimal estimation, that is, the result after high-resolution processing; If the accuracy requirement is not met, repeat step 3 until the accuracy requirement is met.

Images