Longitudinal and transverse wave data matching method based on wavelet compression

Abstract

The invention relates to the technical field of oil and gas geophysical exploration, in particular to a longitudinal and transverse wave data matching method based on wavelet compression, which comprises the following steps: S1, converting and compressing PS wave seismic data to a PP time domain; seismic PP wavelets, PS wavelets, PP wavelet frequency spectrums and PS wavelet frequency spectrums are extracted; s2, introducing a frequency-varying scale transformation factor, and performing wavelet compression on the PS wavelet by using a Fourier transform method; s3, obtaining a filter according to the PS wavelet and the new PS wavelet; s4, filtering the PS wave seismic data after time domain conversion by using a filter; and S5, waveform matching is carried out. On the basis of time domain pre-matching, a frequency-varying scale transformation factor is introduced, and PS wavelets are compressed by using a frequency-varying Fourier scale transformation method. And then a filter is designed by using a regularization method to carry out filtering processing on the original PS wave seismic record, so that the similarity of the waveforms of the PP and PS wave seismic records can be improved, and the final matching precision is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of oil and gas geophysical exploration, and more specifically, to a longitudinal and shear wave data matching method based on wavelet compression. Background Art

[0002] With the development of multi-wave and multi-component seismic technology, multi-wave exploration has become an effective means of reservoir prediction. Compared with single P-wave, the use of multi-wave data for joint inversion and interpretation can effectively improve the accuracy of reservoir prediction. For example, when the target area is a gas-bearing reservoir, the P-wave will be significantly attenuated by the fluid, and a "gas cloud" fuzzy area will appear on the profile, making it difficult to interpret the structure. The propagation of shear waves is mainly related to the stratigraphic skeleton and is almost unaffected by the fluid. Therefore, the interpretation of combined shear waves can help reduce the multi-solution of reservoir prediction. However, in multi-wave exploration, the generation and propagation mechanisms of P-waves and S-waves are different, resulting in large differences in the propagation time and waveform of the collected P-wave and S-wave data. Therefore, accurately matching P-waves with S-waves is the key to joint inversion interpretation, and the quality of the matching results will affect the effect of subsequent joint inversion interpretation.

[0003] In order to combine the information of PP wave (longitudinal wave) and PS wave (converted shear wave), the conventional method is to use the longitudinal and shear wave velocity ratio γ0 to convert and compress the PS wave data into the PP wave time domain, and then perform time shift correction based on the waveform characteristics to match the PS wave and PP wave data. When the accurate longitudinal and shear wave velocity ratio can be obtained, the PS wave reflection time can be corrected to the PP wave reflection time, thereby achieving the purpose of matching. However, in most cases, the longitudinal and shear wave velocity ratio is inaccurate, so some nonlinear matching methods are often used after time domain matching, such as dynamic time warping (DTW), etc., to perform local stretching and compression on the waveform characteristics to improve the matching accuracy.

[0004] However, in actual exploration, the source used is elastic waves. When propagating in the underground medium, they will be affected by the absorption and attenuation of the strata, causing amplitude and frequency band attenuation. And because PP waves and PS waves propagate differently, the degree to which they are affected by stratum absorption and attenuation is also different. Shear waves are a type of shear wave, and there are more external factors from the strata that can cause energy attenuation to them, which makes their amplitude, bandwidth, and phase distortion much greater than longitudinal waves. Generally speaking, in actual seismic data, the bandwidth of converted shear waves is about half of that of longitudinal waves, and the main frequency is also lower than that of longitudinal waves. This difference in the spectrum results in low waveform similarity between the two, longer converted shear wavelets, and lower resolution. Matching the seismic phase axis characteristics on this basis will cause more severe stretching distortion of the converted shear waves, and the final matching results will be inaccurate. Summary of the invention

[0005] In view of the problem of inaccurate P-wave and S-wave matching results in the above-mentioned prior art, the present invention provides a P-wave and S-wave data matching method based on wavelet compression, which can improve the similarity of the seismic recording waveforms of PP and PS waves and improve the accuracy of the final matching.

[0006] In order to solve the above technical problems, the technical solution provided by the present invention is:

[0007] A longitudinal and shear wave data matching method based on wavelet compression comprises the following steps:

[0008] S1: Convert and compress the PS wave seismic data into the PP time domain to obtain the PS wave seismic data after time domain conversion, and realize time domain pre-matching; then extract the seismic PP wavelet, PS wavelet, PP wavelet spectrum and PS wavelet spectrum according to the PP wave seismic data and the PS wave seismic data after time domain conversion.

[0009] S2: Introduce frequency-variable scaling factor a as a compression or stretching scale, use Fourier transform method to perform wavelet compression on PS wavelet, make PS wavelet spectrum consistent with PP wavelet spectrum, and obtain new PS wavelet spectrum. Wherein, the frequency-variable scaling factor changes with the change of frequency. In actual situations, the wavelet spectrum of PP wave and PS wave is not a single stretching or compression relationship, but presents a nonlinear relationship on each frequency component, that is, the fixed scaling factor cannot completely match the two spectra. Therefore, based on the conventional Fourier scaling theory, a frequency-variable scaling factor is proposed, that is, a is no longer a fixed value, but changes with frequency, and the spectrum can be locally stretched and compressed according to the spectral characteristics of the two to improve the matching effect. Under normal circumstances, PP wave has a wide bandwidth, high wavelet resolution and standard waveform, and it contains richer information, so it is only necessary to scale the PS wave to make its spectrum approach the longitudinal wave, compress the wavelet to improve the resolution, and the matching effect is more accurate.

[0010] S3: Obtain a filter based on the PS wavelet and the new PS wavelet. After wavelet compression, the similarity between PS and PP wave seismic recording waveforms will be effectively improved. In order to stabilize the process, a filter is introduced.

[0011] S4: Using the filter to filter the PS wave seismic data after the time domain conversion to obtain new PS wave seismic data.

[0012] S5: Perform waveform matching based on the PP wave seismic data and the new PS wave seismic data.

[0013] Preferably, in step S1, the PS wave seismic data is first converted and compressed into the PP time domain according to the initial aspect ratio velocity ratio γ0 to achieve time domain pre-matching; then the PP wavelet, the PS wavelet, the PP wavelet spectrum and the PS wavelet spectrum are extracted from the PP wave seismic data and the pre-matched PS wave seismic data respectively using a spectral simulation wavelet estimation method.

[0014] Preferably, in step S2, the Fourier transform spectrum is expressed as:

[0015] W(ω)=|W(ω)|e iφ(ω) ;

[0016] Where |W(ω)| is the modulus of the complex number, which is the amplitude spectrum; φ(ω) is the argument of the complex number, which is the phase spectrum; i is the imaginary unit;

[0017] For the seismic wavelet w(t), the relationship between the seismic wavelet and the frequency-varying scale transformation factor is:

[0018]

[0019] Where a is the frequency-varying scaling factor.

[0020] Preferably, in step S2, the frequency-variant scaling factor expression is:

[0021]

[0022] In the formula, ω r is the reference frequency; n>1, m>2, and the values ​​of the two are determined according to the characteristics of the longitudinal wave frequency band. It can be found that when it is in the low frequency band, a<1, and the spectrum is stretched to the low frequency; when it is in the middle frequency band-when the frequency of the converted shear wave and longitudinal wave spectrum overlaps, 1≤a<2, and the spectrum is stretched in a small range; when it is in the high frequency band, a=m, and the spectrum is stretched to the high frequency. Of course, various other situations may occur in actual situations. Assuming that in the low frequency band, the spectra of the two have matched, and there is no need to supplement the low-frequency component of the converted shear wave, a should be equal to 1 at this time. Therefore, when processing actual data, it is necessary to select an appropriate reference frequency ω based on the specific situation. r , parameters m and n. In summary, through the frequency-variable Fourier scaling transform, the converted shear wave spectrum and the longitudinal wave spectrum can be better matched in the frequency domain, and the converted shear wave sub-wave is compressed to make the time domain waveforms of the two consistent, which can improve the matching effect.

[0023] Preferably, in step S2, ω r The value of is half of the main frequency of the longitudinal wave.

[0024] Preferably, in the step S3, a new PS wavelet spectrum is calculated according to the new PS wavelet, and then the filter is obtained according to the relationship between the new PS wavelet spectrum and the wavelet spectrum.

[0025] Order pp (t) is the longitudinal wave seismic record, s ps (t) is the seismic record after the shear wave is converted to the longitudinal wave time domain, and t is the longitudinal wave time. At this time, it is considered that the two time domains are basically corresponding. According to the convolution model, the seismic record can be expressed as the convolution of the seismic wavelet w(t) and the reflection coefficient r(t) in the time domain, and can be expressed as the product of the seismic wavelet spectrum W(ω) and the reflection coefficient spectrum R(ω) in the frequency domain. Then, assuming that the reflection coefficients of the PP wave and the PS wave are consistent, their seismic records can be expressed in the frequency domain as follows:

[0026]

[0027] In the formula, S pp (ω) is the PP wave seismic record spectrum; S ps (ω) is the PS wave seismic record spectrum; W pp (ω) is the PP wavelet spectrum; W ps (ω) is the PS wavelet spectrum.

[0028] Preferably, in step S4, the reflection coefficients of the PP wave and the PS wave are set to be consistent, and the new PS wave seismic record spectrum is expressed as:

[0029]

[0030] In the formula, It is the new PS wave seismic record spectrum; is the new PS wavelet spectrum; R(ω) is the reflection coefficient spectrum; the new PS wavelet spectrum is obtained according to the above formula. The new time domain PS wave seismic record can be obtained by inverse Fourier transform.

[0031] Preferably, in step S4, the new PS wave seismic record spectrum is expressed as:

[0032] In the formula, is the new PS wavelet spectrum; W ps (ω) is the PS wavelet spectrum; S ps (ω) is the PS wave seismic record spectrum; H(ω) is the filter; the filter is obtained according to the above formula. The original seismic record can be directly processed by the filter to obtain a new seismic record. The key to this process is to obtain the filter.

[0033] From the above formula, we can get

[0034] Preferably, in step S4, the filter is obtained by solving using a regularization method.

[0035] If we know the original wavelet spectrum W of the PS wave seismic record ps (ω) and the wavelet spectrum after compression If, preferably, in step S4, the regularized solution vector expression of the filter is:

[0036]

[0037] Where H is the filter vector; W ps is the original PS wavelet spectrum vector, is its complex conjugate representation; is the new PS wavelet spectrum vector; λ is the regularization parameter; I is the identity matrix. Adding λI can avoid the problem of being unable to invert if the matrix is ​​singular.

[0038] The beneficial effects of the present invention are as follows: on the basis of time domain pre-matching, a frequency-variable scaling factor is introduced, and the frequency-variable Fourier scaling method is used to broaden the PS wavelet frequency band, so that it is consistent with the PP wavelet frequency band, thereby playing the role of compressing the PS wavelet. After wavelet compression, the new PS wavelet is basically consistent with the PP wavelet waveform. Then, a filter is designed by using a regularization method to filter the original PS wave seismic record, which can improve the similarity of the PP and PS wave seismic record waveforms and improve the accuracy of the final matching. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a schematic diagram of multiple wave reflections after longitudinal wave incidence;

[0040] Figure 2 It is a flow chart of a method for matching longitudinal and transverse wave data based on wavelet compression;

[0041] Figure 3 It is a schematic diagram of a multi-layer uniform medium model in the depth domain;

[0042] Figure 4 It is a schematic diagram of the synthetic seismic record section of PP wave and PS wave;

[0043] Figure 5 This is a single-channel comparison diagram of PP wave and PS wave synthetic seismic records; the solid line in the figure represents PP wave, and the dotted line represents PS wave;

[0044] Figure 6 It is a schematic diagram of the seismic record profile of PP wave and PS wave after pre-matching in the time domain;

[0045] Figure 7This is a single-channel comparison diagram of PP and PS wave seismic records after time domain pre-matching. The solid line in the figure represents the PP wave, and the dotted line represents the PS wave.

[0046] Figure 8 It is the PP and PS wavelets extracted and their amplitude spectra after time domain pre-matching. The solid line in the figure represents the PP wave, and the dotted line represents the PS wave.

[0047] Fig. 9 It is a schematic diagram of the change of frequency-dependent scaling factor with frequency;

[0048] Fig.10 It is a schematic diagram of PP, new PS wavelet and its amplitude spectrum after wavelet compression;

[0049] Fig.11 It is a schematic diagram of the final PP and PS wave seismic record profile;

[0050] Fig.12 This is a single-channel comparison diagram of the PP wave and the new PS wave seismic record after wavelet compression; the solid line in the figure represents the PP wave, and the dotted line represents the PS wave. DETAILED DESCRIPTION

[0051] The technical solution of the present invention is further described in detail below through specific embodiments and in conjunction with the accompanying drawings:

[0052] Example 1

[0053] Figure 1 The figure shows the phenomenon of converted shear waves when incident longitudinal waves encounter elastic interfaces in seismic exploration. R1 and R2 represent the elastic interfaces of the formation; P waves represent incident longitudinal waves, PP waves represent reflected longitudinal waves, and PS waves represent reflected converted shear waves; V p1 、V s1 、V p2 and V s2 Represent the longitudinal and transverse wave velocities of the upper and lower layers of the medium respectively; α1, β1, α2 and β2 represent the reflection angles of PP and PS waves when the waves propagate in the first and second layers respectively; s pp1 、s ps1 、s pp2 and ps2 The PP and PS waves represent seismic records of waves reflected by one layer of medium and two layers of medium, respectively. Due to the different propagation paths and speeds, theoretically, the response time of the stratum reflection of the PP and PS waves is different, and there is a time shift. Generally, the PS wave is later than the PP wave. In addition, due to the different attenuation effects of the stratum on the longitudinal and transverse waves, the sub-wave waveforms are different, and the amplitude energy of the received PP and PS waves will also be different.

[0054] This embodiment is a first embodiment of a method for matching longitudinal and transverse wave data based on wavelet compression, which can improve the similarity of the seismic recording waveforms of PP and PS waves and improve the accuracy of the final matching. Figure 2 As shown, the method comprises the following steps:

[0055] S1: Convert and compress the PS wave seismic data into the PP time domain to obtain the PS wave seismic data after time domain conversion, and realize time domain pre-matching; then extract the seismic PP wavelet, PS wavelet, PP wavelet spectrum and PS wavelet spectrum according to the PP wave seismic data and the PS wave seismic data after time domain conversion.

[0056] S2: Introduce frequency-variant scaling factor a as a compression or stretching scale, use Fourier transform method to perform wavelet compression on PS wavelet, make the PS wavelet spectrum consistent with the PP wavelet spectrum, and obtain a new PS wavelet spectrum. Among them, the frequency-variant scaling factor changes with the change of frequency. In actual situations, the wavelet spectra of PP wave and PS wave are not a single stretching or compression relationship, but present a nonlinear relationship on each frequency component, that is, the fixed scaling factor cannot completely match the two spectra. Therefore, based on the conventional Fourier scaling theory, a frequency-variant scaling factor is proposed, that is, a is no longer a fixed value, but changes with frequency, which can locally stretch and compress the spectrum according to the spectral characteristics of the two to improve the matching effect. Under normal circumstances, the PP wave has a wide bandwidth, high wavelet resolution and standard waveform, and it contains richer information, so it is only necessary to scale the PS wave to make its spectrum approach the longitudinal wave, compress the wavelet to improve the resolution, and the matching effect is more accurate.

[0057] S3: Obtain the filter based on the PS wavelet and the new PS wavelet. After wavelet compression, the similarity between the PS and PP wave seismic record waveforms will be effectively improved. In order to stabilize the process, a filter is introduced.

[0058] S4: Filter the PS wave seismic data after time domain conversion using a filter to obtain new PS wave seismic data.

[0059] S5: Based on the PP wave seismic data and the new PS wave seismic data, the time domain nonlinear matching methods such as dynamic time warping (DTW) are used to perform waveform matching according to the formation response characteristics to obtain the final multi-wave matching results.

[0060] Furthermore, in step S1, the PS wave seismic data is first converted and compressed into the PP time domain according to the initial aspect ratio velocity ratio γ0 to achieve time domain pre-matching; then the PP wavelet, PS wavelet, PP wavelet spectrum and PS wavelet spectrum are extracted from the PP wave seismic data and the pre-matched PS wave seismic data respectively using the wavelet estimation method of spectral simulation.

[0061] Furthermore, in step S4, a filter is obtained by solving using a regularization method.

[0062] Effect verification of this embodiment: A multi-layer model is designed to verify the feasibility of the method. The model parameters are shown in Table 1. The model has five layers of reflection interfaces, with a depth of about 1200m. The velocity changes with depth as shown in Figure 3 As shown, the P-wave and S-wave velocity ratio is set to 2, and the overall structure presents an anticline structure.

[0063] Table 1 Parameter information of multi-layer model in depth domain

[0064]

[0065] The model is used to synthesize seismic record profiles. The two-way travel time of self-excited and self-received reflections at the stratum interface of PP waves and PS waves is calculated according to the P- and S-wave velocities and stratum thickness. In order to simulate the different absorption and attenuation effects of the underground medium on P- and S-waves, 40Hz and 15Hz Ricker wavelets are used to synthesize seismic records, respectively. The final PP and PS wave synthetic seismic record profiles are obtained as follows: Figure 4 To intuitively analyze the different response characteristics of PP and PS waves to the same stratum, the 5th, 10th and 15th synthetic seismic records are compared, as shown in Figure 5 As shown. Since the purpose is only to verify the effect of wavelet compression on PP and PS waveform matching, the amplitude factor is not considered, so the amplitudes of the two are normalized to make the maximum amplitude consistent. The black solid line represents the PP wave seismic record, and the black dotted line represents the PS wave seismic record. From the comparison of profiles and single channels, it can be seen that for the same stratum reflection, the PS wave speed is slower and the two-way travel time is longer. The time difference increases with the depth.

[0066] Generally, seismic inversion interpretation is performed in the PP time domain. According to the P-wave and S-wave velocity ratio, the PS wave synthetic seismic record is converted and compressed into the PP time domain to obtain the PP and PS wave seismic profiles after time domain pre-matching, such as Figure 6 As shown. At this time, it can be seen from the profile that the reflection time and depth of the same stratum are basically corresponding, and as time is compressed, the vertical resolution of the PS wave seismic profile changes and the phase axis becomes narrower. Still take the 5th, 10th, and 15th seismic records for comparison, as shown Figure 7 As shown, it can be found that although the response time of PP and PS waves to the same formation has matched, there are still differences in the waveforms, and this difference will be more obvious in actual data.

[0067] After pre-matching, the PP and PS wavelets are extracted using the spectrum simulation method. The results are as follows: Figure 7As shown, the black solid line is the PP wavelet and its amplitude spectrum, and the black dotted line is the PS wavelet and its amplitude spectrum. It can be found that the main frequency of the PP wavelet is still 40Hz, while the main frequency of the PS wavelet is about 24Hz, which means that after time domain compression, the PS wavelet becomes narrower and the main frequency is improved to a certain extent. Since the PP and PS wavelets used in the implementation of synthetic data in this patent are both Ricker wavelets, after pre-matching in the time domain, the wavelet morphologies of the two are already close. At this time, the PS waveform can be directly transformed to be consistent with the PP wave through some time domain nonlinear matching methods according to the waveform characteristics to complete the final matching. However, in actual seismic records, there are obvious differences in the waveforms of the two. The PS wave is more affected by absorption attenuation, the wavelet tail is longer, and obvious distortion will occur. If they are directly matched according to the waveform characteristics, the error between the two will become larger and larger. Therefore, when faced with complex situations, the method proposed in this patent stretches and compresses the PS wavelet spectrum in the frequency domain, supplements its low-frequency or high-frequency components to make it consistent with the PP wavelet spectrum. The PS wavelet will be compressed in the time domain, and the wavelet morphology will also change. The similarity between the PP and PS waveforms will be improved, which is a more reasonable choice.

[0068] In view of the inconsistency between PS wavelet and PP wavelet, the frequency-variant Fourier scaling transform method is used to compress the PS wavelet. Figure 8 By comparing the amplitude spectra of , it can be found that the PS wavelet spectrum needs to be compressed at low frequencies and stretched at high frequencies. By calculating the amplitude spectra of the two, the scale transformation factor that varies with frequency is obtained. The result is as follows Fig. 9 As shown. Fig. 9 The frequency-variable scaling factor shown in the figure processes the PS wavelet, and the result is compared with the PP wavelet. Fig.10 As shown in the figure, the black solid line is the PP wavelet and its amplitude spectrum, and the black dotted line is the new PS wavelet and its amplitude spectrum after Fourier scaling. By comparison, it can be found that after Fourier scaling wavelet compression, the new PS wavelet and PP wavelet waveforms completely overlap, and the amplitude spectrum bandwidth is consistent, which proves the feasibility of this method and can well handle the consistency of PS wavelet and PP wavelet.

[0069] After obtaining the new PS wavelet after Fourier scale transform wavelet compression, the filter is solved by regularization method, and the PS wave seismic record is filtered to obtain the seismic record profile after overall wavelet compression. The result is compared with the PP wave seismic record. Fig.11 As shown in the figure, it can be found that the PS wave event axis becomes narrower, the resolution is improved, and the profile characteristics are basically consistent with the PP wave. Fig.12 This is a comparison of single-channel seismic records before and after processing, where the black solid line is the PP wave seismic record and the black dotted line is the new PS wave seismic record. Fig.12It can be found that after wavelet compression, the PS wave seismic record and the PP wave seismic record have a very high matching degree and the waveforms correspond accurately, which proves the correctness of the proposed method and has good application prospects.

[0070] Beneficial effects of this embodiment: On the basis of time domain pre-matching, a frequency-variant scaling factor is introduced, and the frequency-variant Fourier scaling method is used to broaden the PS wavelet frequency band, so that it is consistent with the PP wavelet frequency band, which plays a role in compressing the PS wavelet. After wavelet compression, the new PS wavelet is basically consistent with the PP wavelet waveform. Then, a filter is designed using a regularization method to filter the original PS wave seismic record, which can improve the similarity of the PP and PS wave seismic record waveforms and improve the accuracy of the final matching.

[0071] Example 2

[0072] This embodiment further illustrates step S2 based on embodiment 1.

[0073] Further, in step S2, the Fourier transform spectrum is expressed as:

[0074] W(ω)=|W(ω)|e iφ(ω) ;

[0075] Where |W(ω)| is the modulus of the complex number, which is the amplitude spectrum; φ(ω) is the argument of the complex number, which is the phase spectrum; i is the imaginary unit;

[0076] For the seismic wavelet w(t), the relationship between the seismic wavelet and the frequency-varying scale transformation factor is:

[0077]

[0078] Where a is the frequency-varying scaling factor.

[0079] Furthermore, in step S2, the frequency-variant scaling factor expression is:

[0080]

[0081] In the formula, ω r is the reference frequency; n>1, m>2, and the values ​​of the two are determined according to the characteristics of the longitudinal wave frequency band. It can be found that when it is in the low frequency band, a<1, and the spectrum is stretched to the low frequency; when it is in the middle frequency band-when the frequency of the converted shear wave and longitudinal wave spectrum overlaps, 1≤a<2, and the spectrum is stretched in a small range; when it is in the high frequency band, a=m, and the spectrum is stretched to the high frequency. Of course, various other situations may occur in actual situations. Assuming that in the low frequency band, the spectra of the two have matched, and there is no need to supplement the low-frequency component of the converted shear wave, a should be equal to 1 at this time. Therefore, when processing actual data, it is necessary to select an appropriate reference frequency ω based on the specific situation. r, parameters m and n. In summary, through the frequency-variable Fourier scaling transform, the converted shear wave spectrum and the longitudinal wave spectrum can be better matched in the frequency domain, and the converted shear wave sub-wave is compressed to make the time domain waveforms of the two consistent, which can improve the matching effect.

[0082] Further, in step S2, ω r The value of is half of the main frequency of the longitudinal wave.

[0083] The other features, working principles and beneficial effects of this embodiment are consistent with those of Embodiment 1.

[0084] Example 3

[0085] This embodiment further illustrates steps S3 and S4 based on embodiment 2.

[0086] Order pp (t) is the longitudinal wave seismic record, s ps (t) is the seismic record after the shear wave is converted to the longitudinal wave time domain, and t is the longitudinal wave time. At this time, it is considered that the two time domains are basically corresponding. According to the convolution model, the seismic record can be expressed as the convolution of the seismic wavelet w(t) and the reflection coefficient r(t) in the time domain, and can be expressed as the product of the seismic wavelet spectrum W(ω) and the reflection coefficient spectrum R(ω) in the frequency domain. Then, assuming that the reflection coefficients of the PP wave and the PS wave are consistent, their seismic records can be expressed in the frequency domain as follows:

[0087]

[0088] In the formula, S pp (ω) is the PP wave seismic record spectrum; S ps (ω) is the PS wave seismic record spectrum; W pp (ω) is the PP wavelet spectrum; W ps (ω) is the PS wavelet spectrum.

[0089] Further, in step S4, the reflection coefficients of the PP wave and the PS wave are set to be consistent, and the new PS wave seismic record spectrum is expressed as:

[0090]

[0091] In the formula, It is the new PS wave seismic record spectrum; is the new PS wavelet spectrum; R(ω) is the reflection coefficient spectrum; the new PS wavelet spectrum is obtained according to the above formula. The new time domain PS wave seismic record can be obtained by inverse Fourier transform.

[0092] Further, in step S4, the new PS wave seismic record spectrum is expressed as:

[0093]

[0094] In the formula, is the new PS wavelet spectrum; W ps (ω) is the PS wavelet spectrum; S ps (ω) is the PS wave seismic record spectrum; H(ω) is the filter; the filter is obtained according to the above formula. The original seismic record can be directly processed by the filter to obtain a new seismic record. The key to this process is to obtain the filter. From the above formula, we can get

[0095] If we know the original wavelet spectrum W of the PS wave seismic record ps (ω) and the wavelet spectrum after compression If , further, in step S4, the regularization solution vector expression of the filter is:

[0096]

[0097] Where H is the filter vector; W ps is the original PS wavelet spectrum vector, is its complex conjugate representation; is the new PS wavelet spectrum vector; λ is the regularization parameter; I is the identity matrix. Adding λI can avoid the problem of being unable to invert if the matrix is ​​singular.

[0098] The other features, working principles and beneficial effects of this embodiment are consistent with those of Embodiment 2.

[0099] In the specific contents of the above-mentioned specific implementation methods, the various technical features can be combined in any non-contradictory manner. In order to make the description concise, not all possible combinations of the above-mentioned technical features are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0100] Obviously, the above embodiments of the present invention are merely examples for clearly explaining the present invention, and are not intended to limit the implementation methods of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description, and it is not necessary and impossible to list all the implementation methods here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A longitudinal and transverse wave data matching method based on wavelet compression, characterized in that: The following steps are involved: S1: converting and compressing the PS wave seismic data into the PP time domain to obtain the PS wave seismic data after time domain conversion; then extracting the seismic PP wavelet, PS wavelet, PP wavelet spectrum and PS wavelet spectrum according to the PP wave seismic data and the PS wave seismic data after time domain conversion; S2: Introduce a frequency-variable scaling factor as a compression or stretching scale, and use the Fourier transform method to perform wavelet compression on the PS wavelet, so that the PS wavelet spectrum is consistent with the PP wavelet spectrum, and obtain a new PS wavelet; wherein the frequency-variable scaling factor changes with the change of frequency; S3: Obtain a filter according to the PS wavelet and the new PS wavelet; S4: using the filter to filter the PS wave seismic data after the time domain conversion to obtain new PS wave seismic data; S5: Perform waveform matching based on the PP wave seismic data and the new PS wave seismic data.

2. A longitudinal and shear wave data matching method based on wavelet compression according to claim 1, characterized in that: In the step S1, the PS wave seismic data is first converted and compressed into the PP time domain according to the initial aspect ratio velocity ratio γ0; then the PP wavelet, the PS wavelet, the PP wavelet spectrum and the PS wavelet spectrum are extracted from the PP wave seismic data and the PS wave seismic data after the time domain conversion respectively using the spectral simulation wavelet estimation method.

3. The longitudinal and shear wave data matching method based on wavelet compression according to claim 1 is characterized in that: In step S2, the Fourier transform spectrum is expressed as: W(ω)=|W(ω)|e iφ(ω) ; Where |W(ω)| is the modulus of the complex number, which is the amplitude spectrum; φ(ω) is the argument of the complex number, which is the phase spectrum; i is the imaginary unit; For the seismic wavelet w(t), the relationship between the seismic wavelet and the frequency-varying scale transformation factor is: Where a is the frequency-varying scaling factor.

4. The longitudinal and shear wave data matching method based on wavelet compression according to claim 3 is characterized in that: In step S2, the frequency-variant scaling factor expression is: In the formula, ω r is the reference frequency; n>1, m>2, and their values ​​are determined according to the characteristics of the longitudinal wave frequency band.

5. The method for matching longitudinal and transverse wave data based on wavelet compression according to claim 4, characterized in that: In step S2, ω r The value of is half of the main frequency of the longitudinal wave.

6. A longitudinal and shear wave data matching method based on wavelet compression according to any one of claims 1 to 5, characterized in that: In the step S3, a new PS wavelet spectrum is calculated according to the new PS wavelet, and then the filter is obtained according to the relationship between the new PS wavelet spectrum and the wavelet spectrum.

7. A longitudinal and shear wave data matching method based on wavelet compression according to claim 6, characterized in that: In step S4, the new PS wave seismic record spectrum is expressed in the frequency domain as: In the formula, is the new PS wavelet spectrum; W ps (ω) is the PS wavelet spectrum; S ps (ω) is the PS wave seismic record spectrum; H(ω) is the filter; the filter is obtained according to the above formula.

8. The longitudinal and shear wave data matching method based on wavelet compression according to claim 7 is characterized in that: In step S3, the reflection coefficients of the PP wave and the PS wave are set to be consistent, and the new PS wave seismic record spectrum is expressed as: In the formula, It is the new PS wave seismic record spectrum; is the new PS wavelet spectrum; R(ω) is the reflection coefficient spectrum; the new PS wavelet spectrum is obtained according to the above formula.

9. The longitudinal and shear wave data matching method based on wavelet compression according to claim 8, characterized in that: In the step S4, the filter is obtained by using a regularization method.

10. A longitudinal and shear wave data matching method based on wavelet compression according to claim 9, characterized in that: In step S4, the regularized solution vector expression of the filter is: Where H is the filter vector; W ps is the original PS wavelet spectrum vector, is its complex conjugate representation; is the new PS wavelet spectrum vector; λ is the regularization parameter; I is the identity matrix.

Images