A multi-wavelet inversion method and system based on seismic frequency division technology
Through the multi-wave inversion method of seismic frequency division technology, the matching tracking and construction-oriented filtering technology are used to solve the problems of low resolution and strong uncertainty of the traditional single-wave inversion method, and high resolution and stable Reservoir forecast results.
Patent Information
- Application Number
- CN202210032537.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-01-12
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2042-01-12
AI Technical Summary
The traditional single wave seismic inversion method has low resolution, strong uncertainty, unstable algorithms, difficult to converge, and difficult to meet actual needs.
Using multi-wave inversion method based on seismic frequency division technology, the reservoir sensitive parameters are obtained through well logging, the seismic data is extended, and the initial model is generated. The frequency division constraint inversion is performed using matching tracking frequency division technology and structure-oriented filtering. Use different frequency band information of seismic data.
The vertical and horizontal resolution of the inversion result is improved, the sand body tip extinguishing points are clear, the thin and thick layers are clear, and the algorithm is more stable, meeting actual production needs.
Smart Images

Figure CN114371505B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the field of petroleum geophysical exploration, and in particular to a multi-wavelet inversion method and system based on seismic frequency division technology. Background Art
[0002] Seismic inversion is one of the more effective methods among many reservoir prediction technologies. After decades of development, although it has many shortcomings, in actual production, whether it is closure evaluation or oil and gas reservoir description, it is inseparable from the vertical and horizontal prediction of reservoirs based on seismic inversion. Conventional seismic inversion technology mostly obtains inversion results by extracting a single sub-wave of the main frequency of the seismic data. The low-frequency and high-frequency information of the effective frequency band in the seismic data is not fully utilized, and the obtained inversion results have low resolution and strong uncertainty. The commonly used frequency division inversion method in recent years is actually an inversion method without sub-wave extraction and initial model. In theory, it can achieve a relatively high resolution and reflect the contact relationship of the formations. However, in actual application, the algorithm is unstable and difficult to converge, resulting in inversion results that are difficult to meet the needs. Summary of the invention
[0003] The technical problem to be solved by the present invention is that the results of the traditional single wavelet inversion method have low resolution and strong uncertainty. At the same time, its algorithm is unstable and difficult to converge, resulting in the inversion results being difficult to meet the needs. The purpose is to provide a multi-wavelet inversion method and system based on seismic frequency division technology to solve the above problems.
[0004] The present invention is achieved through the following technical solutions:
[0005] A multi-wavelet inversion method based on seismic frequency division technology, comprising:
[0006] Obtain reservoir sensitive parameters of target formations through logging rock physics analysis;
[0007] Select the corresponding seismic data volume according to reservoir sensitive parameters;
[0008] Perform frequency extension on the seismic data volume to obtain broadband seismic data;
[0009] Obtaining a logging sensitive parameter impedance curve based on reservoir sensitive parameters, and generating an initial model using the logging sensitive parameter impedance curve;
[0010] Based on the initial model, the broadband seismic data is gradually integrated from low frequency to high frequency to perform frequency division constrained inversion and obtain a multi-wavelet inversion data volume.
[0011] In the multi-wavelet inversion method based on seismic frequency division technology, preferably, the seismic data volume is extended to obtain broadband seismic data, including:
[0012] The Q value is inverted by using Gabor transform on the seismic data volume to obtain the Q value;
[0013] Based on the Q value, the seismic data is subjected to inverse Q filtering to obtain broadband seismic data.
[0014] In the above-mentioned multi-wavelet inversion method based on seismic frequency division technology, preferably, it also includes increasing the frequency of broadband seismic data by using a spectrum shaping algorithm to obtain high-resolution broadband seismic data;
[0015] The method of performing frequency division constrained inversion based on the initial model to fuse broadband seismic data from low frequency to high frequency step by step to obtain a multi-wavelet inversion data volume preferably also includes: performing frequency division constrained inversion based on the initial model to fuse broadband seismic data from low frequency to high frequency step by step to obtain a multi-wavelet inversion data volume.
[0016] In the multi-wavelet inversion method based on seismic frequency division technology, preferably, frequency division constrained inversion is performed by a method of stepwise fusion of high-resolution broadband seismic data from low frequency to high frequency based on an initial model, including:
[0017] The seismic data frequency division technology based on matching pursuit is used to divide the high-resolution broadband seismic data to obtain the frequency-divided seismic data;
[0018] Performing structural guidance filtering on the frequency-divided seismic data to improve the signal-to-noise ratio of the frequency-divided seismic data and obtain frequency-divided seismic data with a high signal-to-noise ratio;
[0019] Based on the initial model, the frequency-constrained inversion is performed on the frequency-constrained seismic data with high signal-to-noise ratio by step-by-step fusion from low frequency to high frequency to obtain a multi-wavelet inversion data volume.
[0020] The frequency profile constructed by matching pursuit time-frequency atoms effectively avoids the parallel event illusion caused by the conventional frequency division method based on filtering mechanism. At the same time, the frequency division result has the same Ricker wavelet meaning as the original profile, which is suitable for subsequent inversion work. Therefore, the frequency division constrained inversion adopts the seismic data frequency division technology based on matching pursuit. In order to ensure the signal-to-noise ratio of the data, the frequency division profiles are subjected to structure-guided filtering respectively. The structure-guided filtering can effectively suppress random noise, enhance the consistency and continuity of the event, and maintain the result information of the original seismic image. The constrained inversion is realized by the step-by-step fusion method from low frequency to high frequency, which makes full use of the seismic information of different frequency bands.
[0021] In the multi-wavelet inversion method based on seismic frequency division technology, preferably, the Q value includes an interlayer Q value and a mean Q value; the interlayer Q value is expressed as Q i ; The mean Q value is denoted as Q m ;
[0022] The inter-layer Q value Q i The calculation formula is as follows:
[0023]
[0024] Among them, Q i is the inter-layer Q value, i represents the number of layers, t represents the time, Q -1 R is the root mean square of n layers.
[0025] The mean Q value Q m The calculation formula is as follows:
[0026]
[0027] Among them, Q m is the mean Q value, c∈(0,1], and i represents the number of layers.
[0028] In the multi-wavelet inversion method based on seismic frequency division technology, preferably, the Q value is subjected to inverse Q filtering to obtain time-frequency domain data after inverse Q filtering; wherein the inverse Q filtering is expressed as:
[0029]
[0030] Among them, P(τ,ω) is the wave field, ω is the angular frequency, ω h is the tuning parameter, τ is the travel time, and the obtained wave field P(τ,ω) can be transformed by inverse Gabor transform to obtain the seismic record p(t) processed by inverse Q filtering.
[0031] A multi-wavelet inversion system based on seismic frequency division technology, comprising:
[0032] The reservoir parameter acquisition module is used to obtain reservoir sensitive parameters of the target formation through logging rock physics analysis;
[0033] A parameter screening module is used to select the corresponding seismic data volume according to the reservoir sensitive parameters;
[0034] The frequency extension module performs frequency extension on the seismic data volume to obtain broadband seismic data;
[0035] A model generation module, used to obtain a well logging sensitive parameter impedance curve based on reservoir sensitive parameters, and generate an initial model using the well logging sensitive parameter impedance curve;
[0036] The frequency division constrained inversion module is used to perform frequency division constrained inversion based on the initial model by step-by-step fusion of broadband seismic data from low frequency to high frequency to obtain a multi-wavelet inversion data volume.
[0037] In the above-mentioned multi-wavelet inversion system based on seismic frequency division technology, preferably, the frequency extension module includes:
[0038] A Q value acquisition module is used to perform Q value inversion on the seismic data body using Gabor transformation to obtain the Q value;
[0039] The spectrum spreading seismic data acquisition module is used to perform inverse Q filtering on the seismic data based on the existing Q value to obtain broadband seismic data;
[0040] The spectrum shaping module is used to increase the frequency of broadband seismic data by using the spectrum shaping algorithm to obtain high-resolution broadband seismic data.
[0041] In the above-mentioned multi-wavelet inversion system based on seismic frequency division technology, preferably, the frequency division constrained inversion module includes:
[0042] A frequency-division profile acquisition module is used to divide the high-resolution broadband seismic data by using a seismic data frequency-division technique based on matching pursuit to obtain frequency-division seismic data;
[0043] A guided filter construction module is used to perform guided filter construction processing on the frequency-divided seismic data to improve the signal-to-noise ratio of the frequency-divided seismic data and obtain frequency-divided seismic data with a high signal-to-noise ratio;
[0044] The inversion data volume acquisition module is used to perform frequency division constrained inversion on the frequency division seismic data with high signal-to-noise ratio by step-by-step fusion from low frequency to high frequency to obtain a multi-wavelet inversion data volume.
[0045] Based on the traditional inversion method, the present invention adopts the technical idea of frequency-division constrained inversion and makes full use of the seismic information of different frequency bands of seismic data. Compared with the single main frequency seismic inversion, the seismic information involved is more comprehensive and the frequency band is wider, which is more conducive to obtaining inversion results with clear thin and thick layers and high vertical and horizontal resolution.
[0046] Compared with the prior art, the present invention has the following advantages and beneficial effects:
[0047] 1. The multi-wavelet inversion method and system based on seismic frequency division technology provided by the present invention breaks the traditional single wavelet inversion method. By extracting multiple seismic wavelets for inversion, the relatively low-frequency and relatively high-frequency information in the effective frequency band of the seismic are fully utilized. The inversion results have high vertical and horizontal resolution, clear sand body vanishing points, and obvious thickness. Multi-wavelet inversion solves the problems of low resolution and strong uncertainty of the results of the traditional single wavelet inversion method. At the same time, its algorithm results are relatively stable and can better meet actual production needs.
[0048] 2. The present invention provides a multi-wavelet inversion method and system based on seismic frequency division technology. In order to improve the limitations of the single wavelet inversion method, the matching pursuit frequency division technology is adopted to make full use of the information within the effective frequency band of the seismic data and perform frequency division constrained inversion on the seismic data. BRIEF DESCRIPTION OF THE DRAWINGS
[0049] In order to more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and should not be regarded as limiting the scope. For those of ordinary skill in the art, other relevant drawings can be obtained based on these drawings without creative work. In the drawings:
[0050] Figure 1 This is the overall flow chart of multi-wavelet inversion of the present invention;
[0051] Figure 2 This is a flow chart of the multi-wavelet inversion method of the present invention;
[0052] Figure 3 It is a flow chart of the spectrum extension method of the multi-wavelet inversion method of the present invention;
[0053] Figure 4 It is a flow chart of frequency division constraint inversion of the multi-wavelet inversion method of the present invention;
[0054] Figure 5 is the original seismic profile;
[0055] Figure 6 It is the spectrum diagram corresponding to the original seismic section;
[0056] Figure 7 This is the inverse Q seismic profile of the present invention;
[0057] Figure 8 It is the frequency spectrum diagram corresponding to the inverse Q seismic profile of the present invention;
[0058] Fig. 9 It is the spectrum shaping seismic profile of the present invention;
[0059] Fig.10 The frequency spectrum diagram corresponding to the spectrum shaping seismic profile of the present invention;
[0060] Fig.11 This is the result diagram of sparse pulse inversion;
[0061] Fig.12 This is a diagram of the multi-wavelet inversion result of the present invention. DETAILED DESCRIPTION
[0062] The present invention is applicable to reservoir prediction of thin interbeds and massive media. Based on the traditional inversion method, the present invention adopts the technical idea of frequency division constraint inversion, and makes full use of the seismic information of different frequency bands of seismic data. Compared with the single main frequency seismic inversion, the seismic information involved is more comprehensive and the frequency band is wider, which is more conducive to obtaining inversion results with clear thin and thick layers and high vertical and horizontal resolution.
[0063] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with embodiments and drawings. The exemplary embodiments of the present invention and their description are only used to explain the present invention and are not intended to limit the present invention.
[0064] Example
[0065] In the accompanying drawings, Figure 1 The overall flow chart of multi-wavelet inversion of the embodiment is given; Figure 2 A flow chart of a multi-wavelet inversion method of an embodiment is provided; Figure 3 The spectrum expansion flow chart of multi-wavelet inversion method is given; Figure 4 A flow chart of frequency division constraint inversion of the multi-wavelet inversion method of this embodiment is given; Figure 5 The original seismic profile is given; Figure 6 The frequency spectrum corresponding to the original seismic section is given; Figure 7 The reverse Q seismic profile is given; Figure 8 The frequency spectrum corresponding to the inverse Q seismic profile is given; Fig. 9 The spectrum shaping seismic profile is given; Fig.10 The frequency spectrum corresponding to the spectrum shaping seismic section is given; Fig.11 The sparse pulse inversion result diagram is given; Fig.12 The multi-wavelet inversion result diagram is given.
[0066] This embodiment provides a multi-wavelet inversion method and system based on seismic frequency division technology, such as Figure 2 As shown, the method includes:
[0067] S1: Obtain reservoir sensitive parameters of target formations through logging rock physics analysis;
[0068] S2: Select the corresponding seismic data volume according to reservoir sensitivity parameters;
[0069] S3: Perform frequency extension on the seismic data volume to obtain broadband seismic data;
[0070] S4: obtaining a logging sensitive parameter impedance curve based on the reservoir sensitive parameters, and generating an initial model using the logging sensitive parameter impedance curve;
[0071] S5: Based on the initial model, the broadband seismic data is gradually integrated from low frequency to high frequency to perform frequency division constrained inversion to obtain a multi-wavelet inversion data volume.
[0072] like Figure 3 As shown, the method of performing frequency extension on the seismic data volume to obtain broadband seismic data includes:
[0073] S31: Perform Q value inversion on the seismic data volume using Gabor transform to obtain the Q value;
[0074] Since the underground medium is a viscoelastic heterogeneous medium, when elastic waves propagate in it, it will cause energy attenuation and phase dispersion of seismic waves, change the energy and waveform of seismic waves, and reduce the resolution of seismic records. Therefore, it is particularly important to perform Q attenuation compensation on seismic signals. For reflection seismic data, the conventional Q value analysis method compares seismic wavelets at two different depths (or) times. However, it is difficult to pick up a "clean" wavelet without other wavelet interference and noise interference. Therefore, we use Gabor transform to perform Q value inversion on post-stack seismic records to provide data guarantee for the anti-Q filtering in step S32.
[0075] The following is the principle of Q value inversion:
[0076] For reflection seismic data, the conventional Q-value analysis method is to compare two seismic wavelets at different depths (or times). However, it is difficult to pick up a "clean" wavelet without interference from other wavelets and noise. We use Gabor transform to perform Q-value inversion on post-stack seismic records, using the entire seismic record information.
[0077] For a given seismic trace u(t), we can use Gabor transform to generate a time-varying spectrum U(τ,ω), where τ is the travel time and ω is the angular frequency. When only attenuation is considered, the amplitude of the plane wave is expressed as:
[0078]
[0079] Where A0 is the amplitude when τ=0, and the constant Q is -1 is the equivalent Q value between τ=0 and the current time τ.
[0080] For a given amplitude attenuation metric, we can directly use it to design a gain curve to compensate the amplitude spectrum, and then use the theoretical compensation function to fit the gain curve based on the data to estimate the Q value. Although the design of the gain curve only considers the amplitude effect, for the obtained Q value, the inverse Q filter can perform amplitude compensation and phase correction at the same time.
[0081] The one-dimensional attenuation measurement A(χ) (χ=ωτ) is smoothed by median filtering and then normalized The gain curve based on the data is designed as:
[0082]
[0083] In the formula, Λ d The subscript d in represents the gain function derived directly from the data. The theoretical compensation function is expressed as:
[0084]
[0085] Among them, β(χ,Q)=exp(-χ / 2Q). The final Q estimation is transformed into a minimization problem:
[0086]
[0087] For the multi-layer Q structure model, the equivalent Q-1 value is obtained by the Q analysis method based on the attenuation function and the compensation function, and the inter-layer Q-1 value can be further obtained by the method of obtaining the layer velocity from the root mean square velocity. Assuming that the formation Q model is based on the two-way travel time t i (i=1,…,N) is divided into N layers (n=1,…,N), and the equivalent Q-1 value can be regarded as the root mean square Q of the n layers. -1 R Value, then the Q-1 value of the i-th layer of the formation can be obtained from the relationship, and then the inter-layer Q value can be obtained.
[0088]
[0089] In this project, after using the mean Q value obtained by Gabor analysis, the interior point algorithm is used to invert the interlayer Q value. Assume that the stratum is divided into N layers according to the constant Q value, and the interlayer Q value Q i and the mean Q value Q m The calculation formula is:
[0090]
[0091] Assuming the layer thickness is Δt, we have:
[0092] T m =(n-1)Δt+cΔt (1-7)
[0093] Where cΔt is the partial constant layer thickness Δt, c∈(0,1], so according to the formula:
[0094]
[0095] A linear inversion matrix is constructed, and the inter-layer Q value is inverted from the mean Q value using the interior point algorithm.
[0096] S32: performing inverse Q filtering on the seismic data based on the Q value to obtain broadband seismic data;
[0097] Inverse Q filtering is a technology that compensates for the earth's attenuation effect. It can effectively compensate for the amplitude and phase information of seismic data and effectively broaden the frequency band of seismic data.
[0098] The following is the principle of inverse Q filtering:
[0099] The analytical solution of a one-way wave propagating in a viscoelastic medium is:
[0100] P(z+Δz,ω)=P(z,ω)exp[-jk(ω)Δz] (1-9)
[0101] Where j is the imaginary unit, ω is the angular frequency, and k(ω) is the wave number. The Q effect of the formation is expressed by introducing the complex wave number:
[0102]
[0103] Among them, Q r , v r is the quality factor and phase velocity of the reference frequency, ω h is the tuning parameter, γ=1 / πQ.
[0104] Therefore, the inverse Q filter is expressed as:
[0105]
[0106] When the Q value changes continuously with the travel time, the surface wave field is extended to the time depth τ:
[0107]
[0108] Among them, γ(τ)=1 / πQ(τ), further transforming equation (1.12) to obtain:
[0109]
[0110] Where:
[0111]
[0112] σ 2 is the stability factor, which is consistent with the gain limit G of the actual data. lim (dB) relationship is:
[0113] σ 2 =exp[-(0.23G lim +1.63)] (1-15)
[0114] In order to improve the operational efficiency of inverse Q filtering, Gabor transform is used to perform inverse Q filtering in the time-frequency domain. From equation (1-12), for elastic media:
[0115]
[0116] Then equation (1.12) becomes:
[0117]
[0118] Equation (1-17) is an inverse Q filter equation based on Gabor transform. For wave fields at different time depths, Gabor transform is performed on the surface seismic records, and then inverse Q filtering is performed in the time-frequency domain in combination with the amplitude compensation operator and the phase compensation operator. The obtained wave field P(τ,ω) is transformed using inverse Gabor transform to obtain the seismic record p(t) processed by inverse Q filtering.
[0119] The multi-wavelet inversion method based on seismic frequency division technology preferably further includes:
[0120] S33: The broadband seismic data is frequency-enhanced using a spectrum shaping algorithm to obtain high-resolution broadband seismic data.
[0121] The spectrum shaping algorithm is used to increase the frequency to ensure the high resolution of the seismic data. Through the above steps, the bandwidth of the seismic data can be effectively broadened and the resolution of the seismic data can be improved.
[0122] The following is the principle of spectrum shaping: the complex spectrum is used to extract wavelets, the mixed phase wavelets of pre-stack gathers are accurately extracted, and the anti-wavelets are calculated by the user-given expected wavelets, thereby improving the longitudinal resolution of the gathers.
[0123] The relationship between seismic records, seismic wavelets and reflection coefficient sequences can be expressed by the convolution model as follows:
[0124] x(t)=W(t)*ζ(t)+n(t) (3-1)
[0125] Ignoring the noise term, equation (3-1) is expressed in the frequency domain as:
[0126] x(ω)=w(ω)ξ(ω) (3-2)
[0127] By taking the logarithm of equation (3-2), equation (3-2) can be transformed into a linear system
[0128] lnx(ω)=lnw(ω)+lnξ(ω) (3-3)
[0129] Perform inverse Fourier transform on equation (3-3):
[0130]
[0131] They are called the complex spectrum sequences of x(t), w(t), and ξ(t) respectively.
[0132] Due to the difference in "smoothness" between the wavelet and the reflection coefficient sequence, the repetition spectrum of the wavelet is generally near the origin, while the repetition spectrum of the reflection sequence is far away from the origin. Based on this understanding, designing a low-pass filter in the repetition spectrum can achieve the separation of the wavelet and the reflection coefficient in the repetition spectrum.
[0133] That is, the complex spectrum sequence of the well bypass time domain is low-pass filtered to obtain the complex spectrum sequence of the sub-wave, which is equivalent to intercepting near the zero value to obtain
[0134] right Perform FFT transformation (Fast Fourier Transform) to obtain the complex spectrum in the frequency domain because
[0135]
[0136] available
[0137]
[0138] According to the following relationship, the spectrum w(ω) of the extracted wavelet can be obtained as:
[0139]
[0140] In general, φ w Assign it to 0, and then perform ifft transformation (inverse fast Fourier transform) on w(ω) to obtain the zero-phase sub-wave w0(t) extracted by the well bypass.
[0141] The method of performing step-by-step fusion of broadband seismic data from low frequency to high frequency based on the initial model performs frequency division constrained inversion to obtain a multi-wavelet inversion data volume, including:
[0142] Based on the initial model, the high-resolution broadband seismic data are gradually integrated from low frequency to high frequency to perform frequency-constrained inversion and obtain a multi-wavelet inversion data volume.
[0143] like Figure 4 As shown, the method of stepwise fusion of high-resolution broadband seismic data from low frequency to high frequency based on the initial model performs frequency division constrained inversion, including:
[0144] S51: using a seismic data frequency division technique based on matching pursuit to frequency divide the high-resolution broadband seismic data to obtain frequency-divided seismic data;
[0145] The purpose of frequency division of seismic data is to ensure that the frequency division section can be used for subsequent work, and the frequency division result must have sub-wave meaning. At present, there are two common frequency division technologies in seismic data, bandpass filter frequency division and Morlet wavelet frequency division. Morlet wavelet transform has excellent time-frequency localization properties, which can decompose seismic signals into a series of narrowband signals with central frequencies, and can better separate seismic signals of different scales. It is widely used in seismic frequency division processing. However, the same frequency section obtained by the Morlet wavelet transform method has a tuning effect, which produces illusions such as parallel phase axes.
[0146] The same-frequency profile constructed using matching pursuit time-frequency atoms effectively avoids the parallel event illusion caused by the conventional frequency division method based on filtering mechanism. At the same time, the frequency division result has the same Ricker wavelet meaning as the original profile and is suitable for subsequent inversion work. Therefore, this step adopts the seismic data frequency division technology based on matching pursuit.
[0147] S52: performing structural guidance filtering on the frequency-divided seismic data to improve the signal-to-noise ratio of the frequency-divided seismic data and obtain frequency-divided seismic data with a high signal-to-noise ratio;
[0148] In order to ensure the signal-to-noise ratio of the data, the frequency-divided sections are subjected to structural guidance filtering, which can effectively suppress random noise, enhance the consistency and continuity of the event axis, and maintain the result information of the original seismic image.
[0149] S53: Based on the initial model, the frequency-constrained inversion is performed on the frequency-constrained seismic data with high signal-to-noise ratio by step-by-step fusion from low frequency to high frequency to obtain a multi-wavelet inversion data volume.
[0150] Based on the idea of multi-wavelet inversion, the present invention provides a reasonable technical solution. This technical solution is not the only process to realize multi-wavelet inversion. The frequency division in the frequency division constraint inversion method can be replaced by other frequency division methods with wavelet significance. At the same time, the seismic inversion method can also be replaced by other methods such as fusion inversion.
[0151] Take some actual data as an example:
[0152] The reservoir in this work area is relatively thin, with strong lateral heterogeneity, and the reservoir above the well has low resistivity, so impedance inversion is applicable. Conventional seismic inversion methods have low resolution and cannot identify reservoirs, so multi-wavelet high-resolution inversion methods are used.
[0153] like Figure 5 The original seismic data is shown in the figure. The main frequency of the seismic data is 20 Hz, the frequency band is narrow, and the low resolution is the main contradiction. Therefore, the multi-channel anti-Q filtering + frequency boosting method is used to broaden the effective frequency band range of the original seismic data.
[0154] like Figure 7 As shown, compared Figure 5 It can be seen that after the inverse Q filter, the amplitude consistency of the seismic data becomes better. On this basis, spectrum shaping and frequency enhancement are performed, such as Fig. 9 As shown in the figure, after the inverse Q+ spectrum shaping and frequency boosting, the resolution of the seismic data is improved while the lateral continuity of the seismic profile is maintained.
[0155] For comparison, conventional seismic inversion and multi-wavelet inversion were performed, e.g. Fig.11 The results of conventional seismic inversion are shown. Fig.12The results of multi-wavelet inversion are compared and found that the multi-wavelet inversion method not only maintains the same trend as the traditional inversion method, but its resolution is more than double that of the traditional inversion method, which is effective for reservoir prediction in this work area.
[0156] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.
Claims
1. A multi-wavelet inversion method based on seismic frequency division technology, characterized in that: include: Obtain reservoir sensitive parameters of target formations through logging rock physics analysis; Select the corresponding seismic data volume according to reservoir sensitive parameters; Perform frequency extension on the seismic data volume to obtain broadband seismic data; Obtaining a logging sensitive parameter impedance curve based on reservoir sensitive parameters, and generating an initial model using the logging sensitive parameter impedance curve; Based on the initial model, the broadband seismic data is gradually integrated from low frequency to high frequency to perform frequency division constrained inversion to obtain a multi-wavelet inversion data volume; The method of performing frequency extension on the seismic data volume to obtain broadband seismic data includes: performing Q value inversion on the seismic data volume using Gabor transform to obtain the Q value; performing inverse Q filtering on the seismic data based on the Q value to obtain broadband seismic data; The Q value includes the inter-layer Q value and the average Q value; the inter-layer Q value is expressed as Q i ; The mean Q value is denoted as Q m ; The inter-layer Q value Q i The calculation formula is as follows: Among them, Q i is the inter-layer Q value, i represents the number of layers, t represents the time, Q -1 R is the root mean square of n layers; The mean Q value Q m The calculation formula is as follows: Among them, Q m is the mean Q value, c∈(0,1], i represents the number of layers; The inverse Q filtering is performed on the seismic data based on the Q value to obtain broadband seismic data; wherein the inverse Q filtering is expressed as: Among them, P(τ,ω) is the wave field, ω is the angular frequency, ω h is the tuning parameter, τ is the travel time, and the obtained wave field P(τ,ω) can be transformed by inverse Gabor transform to obtain the seismic record p(t) processed by inverse Q filtering.
2. The multi-wavelet inversion method based on seismic frequency division technology according to claim 1, characterized in that: It also includes using a spectrum shaping algorithm to increase the frequency of broadband seismic data to obtain high-resolution broadband seismic data; The method of performing step-by-step fusion of broadband seismic data from low frequency to high frequency based on the initial model performs frequency division constrained inversion to obtain a multi-wavelet inversion data volume, including: Based on the initial model, the high-resolution broadband seismic data are gradually integrated from low frequency to high frequency to perform frequency-constrained inversion and obtain a multi-wavelet inversion data volume.
3. The multi-wavelet inversion method based on seismic frequency division technology according to claim 2, characterized in that: The method of performing step-by-step fusion of high-resolution broadband seismic data from low frequency to high frequency based on the initial model to perform frequency division constrained inversion includes: The seismic data frequency division technology based on matching pursuit is used to divide the high-resolution broadband seismic data to obtain the frequency-divided seismic data; Performing structural guidance filtering on the frequency-divided seismic data to improve the signal-to-noise ratio of the frequency-divided seismic data and obtain frequency-divided seismic data with a high signal-to-noise ratio; Based on the initial model, the frequency-constrained inversion is performed on the frequency-constrained seismic data with high signal-to-noise ratio by step-by-step fusion from low frequency to high frequency to obtain a multi-wavelet inversion data volume.
4. A multi-wavelet inversion system based on seismic frequency division technology, characterized in that: include: The reservoir parameter acquisition module is used to obtain reservoir sensitive parameters of the target formation through logging rock physics analysis; A parameter screening module is used to select the corresponding seismic data volume according to the reservoir sensitive parameters; The frequency extension module performs frequency extension on the seismic data volume to obtain broadband seismic data; A model generation module, used to obtain a well logging sensitive parameter impedance curve based on reservoir sensitive parameters, and generate an initial model using the well logging sensitive parameter impedance curve; The frequency-constrained inversion module is used to perform frequency-constrained inversion based on the initial model by fusing broadband seismic data from low frequency to high frequency step by step to obtain a multi-wavelet inversion data volume; The spectrum extension module includes: a Q value acquisition module, which is used to perform Q value inversion on the seismic data body using Gabor transform to obtain the Q value; a spectrum extension seismic data acquisition module, which is used to perform inverse Q filtering on the seismic data based on the existing Q value to obtain broadband seismic data; a spectrum shaping module, which is used to increase the frequency of broadband seismic data using a spectrum shaping algorithm to obtain high-resolution broadband seismic data; The Q value includes the inter-layer Q value and the average Q value; the inter-layer Q value is expressed as Q i ; The mean Q value is denoted as Q m ; The inter-layer Q value Q i The calculation formula is as follows: Among them, Q i is the inter-layer Q value, i represents the number of layers, t represents the time, Q -1 R is the root mean square of n layers; The mean Q value Q m The calculation formula is as follows: Among them, Q m is the mean Q value, c∈(0,1], i represents the number of layers; Based on the Q value, the seismic data is subjected to inverse Q filtering to obtain broadband seismic data; wherein, the inverse Q filtering is expressed as: Among them, P(τ,ω) is the wave field, ω is the angular frequency, ω h is the tuning parameter, τ is the travel time, and the obtained wave field P(τ,ω) can be transformed by inverse Gabor transform to obtain the seismic record p(t) processed by inverse Q filtering.
5. The multi-wavelet inversion system based on seismic frequency division technology according to claim 4, characterized in that: The frequency division constrained inversion module comprises: A frequency-division profile acquisition module is used to divide the high-resolution broadband seismic data by using a seismic data frequency-division technique based on matching pursuit to obtain frequency-division seismic data; A guided filter construction module is used to perform guided filter construction processing on the frequency-divided seismic data to improve the signal-to-noise ratio of the frequency-divided seismic data and obtain frequency-divided seismic data with a high signal-to-noise ratio; The inversion data volume acquisition module is used to perform frequency-constrained inversion on the frequency-divided seismic data with high signal-to-noise ratio by step-by-step fusion from low frequency to high frequency to obtain a multi-wavelet inversion data volume.
