Seismic waveform classification and discrimination sedimentary microfacies interpretation method

Abstract

The invention relates to the technical field of oil and gas field exploration and development, in particular to a seismic waveform classification and discrimination sedimentary microfacies interpretation method, which comprises the following steps of: S1, optimizing seismic data, and optimizing resolution and quality of the seismic data by adopting a random noise suppression optimization treatment measure; s2, three-transient attribute body calculation and optimization are carried out, and earthquake three-transient attribute body calculation and selection are carried out on the basis of the step S1; s3, seismic waveform clustering optimization: performing optimization selection on a seismic time window, a waveform clustering number and a waveform clustering device type to realize seismic waveform clustering; s4, generating a seismic facies diagram, and generating the seismic facies diagram according to the seismic waveform clustering result; and S5, manufacturing a sedimentary microfacies diagram. According to the method, a geophysical method is taken as a basic research means, a seismic waveform classification and discrimination sedimentary microfacies interpretation method is formed, the method is applied to the full life cycle of exploration and development, the geological problem of exploration and development is solved, and oil and gas field exploration efficiency improvement implementation is supported.

Description

Technical Field

[0001] The invention relates to the technical field of oil and gas field exploration and development, and is a method for classifying and distinguishing sedimentary microfacies interpretation by seismic waveforms. Background Art

[0002] Since the emergence of seismic stratigraphy in the field of oil seismic exploration in the late 1970s, with the development of science and technology and the continuous deepening of data research, the concept of seismic facies has gradually been recognized and widely used. "Facies" refers to the sum of the paleogeographic environment and its material performance when a certain rock layer was formed (Sloss, 1962). Seismic facies analysis is a powerful method for sedimentary facies research. The traditional seismic facies division method is to describe it through naked eye observation, which has great subjectivity and uncertainty. With the development of seismic data processing technology and computer technology, the more popular seismic facies analysis methods at home and abroad mainly include seismic attribute feature mapping method (Gao, 2008) originated from seismic stratigraphy, and seismic geomorphology imaging method (Zeng, 2004) mainly borrowed from the method of geomorphic imaging. In 1984, G. Sibille of France formally proposed the basic principles and techniques of waveform classification attribute analysis and applied it to seismic facies research. In 1999, the French company Paladin applied this technology to commercial software, and this technology has been gradually promoted in China. Summary of the invention

[0003] The present invention provides a method for classifying and distinguishing sedimentary microfacies by seismic waveforms, which overcomes the shortcomings of the above-mentioned prior art and provides a method for using seismic waveforms for sedimentary microfacies interpretation.

[0004] The technical solution of the present invention is achieved by the following measures: A method for classifying and distinguishing sedimentary microfacies by seismic waveforms is carried out according to the following steps:

[0005] S1, optimization processing of seismic data, taking random noise suppression optimization processing measures to achieve optimization of seismic data resolution and quality;

[0006] S2, calculation and optimization of three-moment attribute bodies, based on step S1, calculation and selection of seismic three-moment attribute bodies are performed, and the selected attribute body must conform to the knowledge of geological sedimentation, otherwise, steps S1 and S2 are repeated;

[0007] S3, seismic waveform clustering optimization, optimizes the seismic time window, the number of waveform clusters, and the type of waveform clusterer to achieve seismic waveform clustering;

[0008] S4, generating a seismic phase map, generating a seismic phase map according to the seismic waveform clustering result, analyzing whether the seismic phase is consistent with the geological knowledge on the generated seismic phase map, and generating a map output if it is consistent, otherwise repeating steps S1 and S2;

[0009] S5, prepare sedimentary microfacies map, obtain sedimentary microfacies change information based on seismic facies map, combine single well sedimentary microfacies, and use mapping software to prepare sedimentary microfacies map.

[0010] The following are further optimizations and / or improvements to the above technical solutions:

[0011] In the above step S1, the seismic data is processed by using a random noise suppression optimization processing method. The random noise suppression of the seismic data includes frequency and space. The three-dimensional seismic data is transformed into the frequency-space domain by Fourier transform. The specific operation is:

[0012] S11, for each frequency f component S(i,j,k), a prediction operator U(l,j,k) is designed using the least squares principle, and its prediction error energy Q(f) is expressed as formula 1:

[0013]

[0014]

[0015] in, is the conjugate of V(i,j,k),

[0016] L, M, and N are the lengths of the longitudinal survey line, transverse survey line, and shot offset of the three-dimensional seismic data space window, respectively, and U (l, m, n) is the three-dimensional space window length prediction operator;

[0017] S12, by taking partial derivatives of equations 1 and 2, we get the matrix equation:

[0018] H·U=R Formula 3

[0019] Where: H and R are the covariance matrices from 3D seismic data, and U is the 3D prediction operator;

[0020] S13, for each frequency component, the prediction operator is obtained by solving the matrix equation of formula 3, and then the seismic data is filtered by the prediction operator, and finally converted to the time-space domain by inverse Fourier transform to obtain the result of seismic data denoising.

[0021] In the above step S2, the earthquake three-moment attribute body is calculated by obtaining the relationship between the earthquake information and the geological information over time. The specific operation is:

[0022] S21, decomposes the seismic wavelet into three items: frequency, phase and amplitude, and its convolution model is:

[0023] X(t)=S(t)*ε(t) Formula 4

[0024] Among them, X(t) is the seismic data, s(t) is the excitation wavelet, and ε(t) is the reflection coefficient;

[0025] S22, when the reflection coefficient is convolved with the seismic wavelet, the result is described by the three-moment attributes calculated from the seismic data, namely:

[0026] Instantaneous frequency:

[0027] Instantaneous phase: θ(t) = arctan[x r (t) / x i (t)] Equation 6

[0028] Instantaneous amplitude:

[0029] Among them, x r (t) is the real part of the seismic data channel, x i (t) is the imaginary seismic data channel, θ(t) is the instantaneous phase, f(t) is the instantaneous frequency, and A(t) is the instantaneous amplitude.

[0030] In the above step S3, the selection range of the seismic time window is 50 ms below the bottom interface of the target layer to 50 ms above the top interface.

[0031] In the above step S3, the number of waveform clusters is 5 to 8.

[0032] In the above step S3, the waveform clusterer type is a supervised waveform classifier or an unsupervised clusterer. The supervised waveform classifier is applicable to the stage with a higher degree of well control, and the unsupervised waveform classifier is applicable to the stage with a lower degree of well control.

[0033] In the above step S4, when generating the seismic phase map, the seismic target layer is divided into different seismic phase zones by utilizing the differences in the characteristics of seismic waveform parameters, and then the sedimentary environment is inferred to generate the seismic phase map.

[0034] The present invention takes geophysics, sedimentology, and stratigraphic sequence science as the basic theoretical system, and geophysical methods as the basic research means, to form the method results and understanding of seismic waveform classification and sedimentary microfacies interpretation, which are applied to the entire life cycle of exploration and development, solve geological problems of exploration and development, and support the implementation of quality improvement and efficiency enhancement in oil and gas field exploration. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Attached Figure 1 The present invention is a flow chart of the method.

[0036] Attached Figure 2 This is a comparative analysis diagram of the time profile before and after random noise suppression in Example 8.

[0037] Attached Figure 3This is a diagram of extraction and comparative analysis of the three instantaneous earthquake (amplitude, frequency, phase) attributes in Example 8.

[0038] Attached Figure 4 A comparative analysis diagram is selected for the earthquake time window in Example 8.

[0039] Attached Figure 5 This is a comparative analysis diagram for selecting the waveform clustering number in Example 8.

[0040] Attached Figure 6 A comparative analysis graph is selected for the supervised waveform classifier and the unsupervised waveform classifier in Example 8.

[0041] Attached Figure 7 This is a comparative plan view of the T1b2 depositional microstructure of the Madong slope in Example 8. DETAILED DESCRIPTION

[0042] The present invention is not limited by the following embodiments, and specific implementation methods can be determined based on the technical solution of the present invention and actual conditions.

[0043] The present invention will be further described below in conjunction with embodiments:

[0044] Embodiment 1: The method for interpreting sedimentary microfacies by classification of seismic waveforms is carried out according to the following steps:

[0045] S1, optimization processing of seismic data, taking random noise suppression optimization processing measures to achieve optimization of seismic data resolution and quality;

[0046] S2, calculation and optimization of three-moment attribute bodies, based on step S1, calculation and selection of seismic three-moment attribute bodies are performed, and the selected attribute body must conform to the knowledge of geological sedimentation, otherwise, steps S1 and S2 are repeated;

[0047] S3, seismic waveform clustering optimization, optimizes the seismic time window, the number of waveform clusters, and the type of waveform clusterer to achieve seismic waveform clustering;

[0048] S4, generating a seismic phase map, generating a seismic phase map according to the seismic waveform clustering result, analyzing whether the seismic phase is consistent with the geological knowledge on the generated seismic phase map, and generating a map output if it is consistent, otherwise repeating steps S1 and S2;

[0049] S5, prepare sedimentary microfacies map, obtain sedimentary microfacies change information based on seismic facies map, combine single well sedimentary microfacies, and use mapping software to prepare sedimentary microfacies map.

[0050] The present invention is based on the comprehensive manifestation of geological information and geophysical information contained in underground geological bodies by seismic waves. Since different types of sedimentary phases usually have different seismic phase characteristics, by analyzing the reflection intensity, waveform changes, and superposition relationship of each phase axis that constitutes a certain type of seismic phase, the internal structure and spatial distribution of the sedimentary body can be finely portrayed, thereby guiding the interpretation of sedimentary microfacies and providing a reliable geological basis for development and potential tapping.

[0051] The method for interpreting sedimentary microfacies by seismic waveform classification provided by the present invention can make full use of the information-rich characteristics of seismic data, optimize the processing of seismic data, calculate and optimize the three-moment attribute bodies, and generate a seismic phase map to obtain sedimentary microfacies change information through the above steps and repeated quality control and parameter (time window selection, waveform clustering number, classifier selection) testing. On this basis, according to the seismic phase map, combined with the single well sedimentary microfacies, the sedimentary microfacies map is drawn using CorelDRAW mapping software, which intuitively describes and reveals the sedimentary environment of the study area, providing important spatial information for oilfield exploration and development.

[0052] Embodiment 2: As an optimization of the above embodiment, in step S1, the seismic data is processed by using a random noise suppression optimization processing method. The random noise suppression of the seismic data includes frequency and space. The three-dimensional seismic data is transformed into the frequency-space domain by Fourier transform. The specific operations are as follows:

[0053] S11, for each frequency f component S(i,j,k), a prediction operator U(l,j,k) is designed using the least squares principle, and its prediction error energy Q(f) is expressed as formula 1:

[0054]

[0055]

[0056] in, is the conjugate of V(i,j,k),

[0057] M, M, and N are the lengths of the longitudinal survey line, transverse survey line, and shot offset of the three-dimensional seismic data space window, respectively, and U(l,m,n) is the three-dimensional space window length prediction operator;

[0058] S12, by taking partial derivatives of equations 1 and 2, we get the matrix equation:

[0059] H·U=R Formula 3

[0060] Where: H and R are the covariance matrices from 3D seismic data, and U is the 3D prediction operator;

[0061] S13, for each frequency component, the prediction operator is obtained by solving the matrix equation of formula 3, and then the seismic data is filtered by the prediction operator, and finally converted to the time-space domain by inverse Fourier transform to obtain the result of seismic data denoising.

[0062] Seismic data is the basis of waveform classification. It is necessary to strictly control the seismic data, understand the quality and problems of the data, and make targeted corrections to ensure the reliability of waveform classification. After optimization, the consistency of seismic data has improved, the abnormal boundaries of amplitude attributes are clear, the details of amplitude attributes are rich, and the overall quality is improved.

[0063] Embodiment 3: As an optimization of the above embodiment, in step S2, the earthquake three-moment attribute body is calculated by obtaining the relationship between the earthquake information and the geological information over time. The specific operation is as follows:

[0064] S21, decomposes the seismic wavelet into three items: frequency, phase and amplitude, and its convolution model is:

[0065] X(t)=S(t)*ε(t) Formula 4

[0066] Among them, X(t) is the seismic data, s(t) is the excitation wavelet, and ε(t) is the reflection coefficient;

[0067] S22, when the reflection coefficient is convolved with the seismic wavelet, the result is described by the three-moment attributes calculated from the seismic data, namely:

[0068] Instantaneous frequency:

[0069] Instantaneous phase: θ(t) = arctan[x r (t) / x i (t)] Equation 6

[0070] Instantaneous amplitude:

[0071] Among them, x r (t) is the real part of the seismic data channel, x i (t) is the imaginary seismic data channel, θ(t) is the instantaneous phase, f(t) is the instantaneous frequency, and A(t) is the instantaneous amplitude.

[0072] The present invention determines the attribute body that can accurately reflect the important information of favorable reservoir phase belts for seismic waveform clustering by studying the direct or indirect relationship between the three seismic attributes and sedimentary microfacies, and calculating and optimizing the three seismic attributes. The three seismic attributes are instantaneous frequency, instantaneous phase, and instantaneous amplitude. The relationship between the three seismic attributes and sedimentary microfacies is:

[0073] ① The instantaneous frequency is related to the natural frequency of the layer from which information is extracted from seismic data, and the natural frequency of the layer is related to the coarseness of the sediment particles. From the perspective of resonance, the resonance frequency is low when the sediment particles are coarse, and high when the sediment particles are fine.

[0074] ② The instantaneous phase is related to the viscosity of the strata from which information is extracted from seismic data to the main frequency of the earthquake. That is, when seismic waves pass through strata of different lithologies, they will cause changes in the phase of the seismic waves. Therefore, this attribute is more sensitive in describing the boundaries of sedimentary microfacies.

[0075] ③The instantaneous amplitude is related to the formation reflection coefficient extracted from seismic data (related to the velocity, density, and pore fluid properties of the formation), so it has a direct relationship with the sedimentary microfacies.

[0076] Embodiment 4: As an optimization of the above embodiment, in step S3, the selection range of the seismic time window is 50 ms below the bottom interface of the target layer to 50 ms above the top interface.

[0077] The seismic time window is a collection of seismic data between two horizons or a certain horizon plus a lower limit range. Since the seismic waveform comes from the target geological body, the selection of the time window is very critical. If the selected time window is too small, on the one hand, the geological target will be beyond the extraction range, and on the other hand, a too small time window will affect the validity of certain parameters; and if the selected time window is too large, the characteristic anomalies of the geological target will be submerged in other background anomalies. In order to select a suitable time window, according to the sedimentary type of the reservoir, the present invention selects 50ms from the bottom interface of the target layer to the top interface as the seismic time window range for testing until the sedimentary microfacies can be finely portrayed. In short, before selecting the seismic time window, it is necessary to understand the approximate depth, thickness, and velocity of the geological body in order to accurately determine the size of the extraction range of the target layer time window.

[0078] Embodiment 5: As an optimization of the above embodiment, in step S3, the number of waveform clusters is 5 to 8.

[0079] The number of seismic waveform clusters refers to the number of types of seismic traces encountered in the target layer. In actual operation, it is difficult to define a universal ideal cluster number. The correct number of waveform clusters should depend on the research objectives and the degree of understanding of the data. If the number of waveform clusters is large, the results are too detailed, and if the number of waveform clusters is small, the results are too rough. It is usually difficult to interpret more than 15 or 20 types. Selecting 5 to 8 waveform clusters has a better matching effect with the sedimentary microfacies type.

[0080] Embodiment 6: As an optimization of the above embodiment, in step S3, the waveform clusterer type is a supervised waveform classifier or an unsupervised clusterer. The supervised waveform classifier is applicable to the stage with a high degree of well control (the development stage has a clear understanding of the geology of the study area), and the unsupervised waveform classifier is applicable to the stage with a low degree of well control (exploration stage). The waveform clusterer can automatically classify the seismic waveforms in the time window and generate a waveform classification map.

[0081] The supervised waveform classifier requires the user to specify the location of the reference waveform (the sedimentary microphase corresponding waveform within the seismic time window at the well location or the sedimentary microphase corresponding waveform within the seismic time window determined by the line number). The best control method is to specify the reference waveform at the well location in combination with previous geological knowledge. The principle of supervised classification is to compare all traces in the data body with the user-defined waveform along the seismic time window, and the relevant calculation uses the Euclidean distance similarity calculation technology.

[0082] The principle is: Assume that the given data set X = {x m |m=1,2,…,total}, the sample in the formula is described by d attributes A 1 , A 2 , …A d To represent, and the d descriptive attributes are all continuous attributes. Data sample X i =(X i1 , X i2 ,…X id ), X j =(X j1 , X j2 ,…X jd ). Among them, X i1 ,X i2 ,…X id and X j1 , X j2 ,…X jd They are samples X i and X j Corresponding to d descriptive attributes A 1 , A 2 , …A d The specific value of . Sample X i and X j The similarity between them is usually expressed as the distance d(X i ,X j ) to indicate that the smaller the distance, the closer the sample X i and X j The more similar, the smaller the difference; the larger the distance, the more sample X i and X j The less similar, the greater the difference. The Euclidean distance formula is as follows:

[0083]

[0084] In the formula, q i is the reference waveform; p i is the target waveform on the layer; N is the waveform time sampling point.

[0085] Unsupervised clusterer, whose algorithm is fast waveform clustering, does not require the user to specify a reference waveform, and the groups are roughly sorted by similarity. Multiple groups are defined, each group represents a typical (reference waveform) waveform found in the data, and the group numbers are sorted by similarity. The correlation calculation uses the Manhattan distance similarity calculation technique.

[0086] Manhattan distance or City Block is an effective statistical measure of similarity or dissimilarity. Manhattan distance uses two waveforms of equal length, N time samples, and sums the absolute values ​​of the sample differences corresponding to all samples. Therefore, the Manhattan distance calculation formula is:

[0087]

[0088] Where M is the Manhattan distance, A is the reference waveform, B is the target waveform on the layer, and N is the waveform time sampling point.

[0089] In summary, the use of supervised clusterers requires the user to specify one or more reference waveforms, and waveform clustering is performed by calculating similarity through Euclidean distance; unsupervised clusterers are a type of fast waveform clustering that does not require the user to specify reference waveforms, uses Manhattan distance to calculate similarity, and clusters waveforms by sorting by group number.

[0090] Embodiment 7: As an optimization of the above embodiment, in step S4, when generating a seismic phase map, the seismic target layer is divided into different seismic phase zones by utilizing the differences in the characteristics of seismic waveform parameters, and then the sedimentary environment is inferred to generate a seismic phase map.

[0091] The method for interpreting sedimentary microfacies by classifying and distinguishing seismic waveforms of the present invention uses seismic waveforms with the same characteristics as a complex to conduct seismic phase research and interpret sedimentary microfacies, which is based on the comprehensive manifestation of geological information and geophysical information contained in the underground geological body by seismic waves. Since different types of sedimentary phases usually have different seismic phase characteristics, the internal structure and spatial distribution of the sedimentary body can be finely portrayed by analyzing the reflection intensity, waveform changes, and superposition relationship of each phase axis constituting a certain type of seismic phase, thereby guiding the interpretation of sedimentary microfacies and providing a reliable geological basis for development and potential tapping.

[0092] Example 8: The Baikouquan Formation in the Da 13 well area of ​​the Mahu Sag was taken as the research object, and the method of the present invention for interpreting sedimentary microfacies by seismic waveform classification was used for interpretation. The specific application steps are as follows:

[0093] Step S1, optimization processing of seismic data. Establish an optimization processing flow for seismic data, and carry out targeted corrections on the quality and existing problems of the original seismic data, effectively solving the problems of low signal-to-noise ratio and low resolution of the original data. Figure 2 Comparison of seismic data before and after processing in the Da 13 well area shows that seismic data after random noise suppression and optimization can more effectively depict geological features such as folds and landforms, reducing the multiple solutions of sedimentary phase identification caused by low signal-to-noise ratio and low resolution of seismic data. This also verifies the rationality of this seismic data optimization processing.

[0094] Step S2, calculation and optimization of three-moment attribute bodies. The three-moment attribute bodies of the seismic are extracted from the T1b2 layer of the Baikouquan Formation in the Da13 well area of ​​the Mahu Sag for 50ms each, and then analyzed and optimized. Figure 3 It can be seen that the imaging details of frequency attributes are not enough, and the amplitude and phase attributes can clearly reflect the phase and amplitude changes of the T1b2 layer; the phase attribute clearly depicts the sedimentary microfacies boundary of the T1b2 layer, and the amplitude attribute is second. Therefore, the phase attribute body is selected for seismic waveform clustering.

[0095] Step S3, seismic waveform clustering optimization. Seismic waveform clustering performs clustering and research and analysis on the phase attribute body, including the determination of the seismic time window, the selection of the number of clusters, and the selection of the supervised clusterer.

[0096] ① Selection and determination of seismic time window. In order to select a suitable time window, according to the sedimentary type of the reservoir, the time window range is selected from the lower 50ms of the target layer interface to the upper 50ms of the top interface, and the sedimentary microfacies can be finely portrayed. Comparative analysis of 10ms, 20ms, 30ms above the T1b2 layer to the lower 10ms, 20ms, 30ms and 5ms above the T1b2 layer and 5ms above the T1b2 layer shows that the time window range of 5ms above the T1b2 layer in this study area is consistent with the sedimentary understanding.

[0097] ② Selection of waveform clustering number. After repeated experiments and combined with the calibration of single well phase, it is determined that 8 types of seismic waveform clustering numbers can well reflect the actual sedimentary microfacies type. Figure 5 It can be seen from the seismic phase diagrams obtained by selecting cluster numbers 5, 8, and 10 that the seismic phases when the cluster number is 5 cannot clearly show the sedimentary environment, the seismic phase divisions when the cluster number is 10 are messy, and when the cluster number is 8, the seismic phases clearly depict the sedimentary microfacies of this study area.

[0098] ③Selection of waveform clusterer. Figure 6 In this paper, by comparing and analyzing the seismic facies maps of supervised and unsupervised waveform classifiers, the unsupervised waveform classifier can accurately characterize the sedimentary microfacies of this area.

[0099] Step S4, making a sedimentary microfacies map. In this example, pure wave data is selected for optimization processing to ensure the authenticity of the information; the time window range of 5ms upward from the target layer is determined; for the research objectives, waveform clustering is performed into 8 categories and unsupervised waveform classifier is used to perform waveform clustering analysis to generate the T1b2 seismic phase plane map of the Madong slope. Figure 7 In the figure, the sedimentary microfacies map of T1b2 of the Madong slope (right) was generated after the sedimentary microfacies were identified by the seismic facies map (left). From the comparison of the two pictures, it can be seen that the consistency is high. The sedimentary microfacies plane map reflects the distribution area of ​​favorable sand bodies in the study area and points out the next favorable exploration and development direction.

[0100] The present invention produces sedimentary microfacies maps through data processing and optimization (optimization of seismic data, calculation and optimization of three-moment attribute bodies), seismic waveform clustering analysis (selection and determination of seismic time windows, selection of waveform clustering numbers, selection of supervised waveform classifiers or unsupervised waveform classifiers), and three major categories and five steps for seismic waveform classification and interpretation of sedimentary microfacies. A detailed comparative study was conducted on the application conditions of the seismic waveform classification method, and on this basis, the method of seismic waveform classification and interpretation of sedimentary microfacies was applied to the study of the Baikouquan Formation in the Da 13 well area of ​​the Mahu Sag, and a good application effect was achieved.

[0101] In summary, the present invention provides a method for interpreting sedimentary microfacies by classifying seismic waveforms. It takes geophysics, sedimentology, and stratigraphic sequence as the basic theoretical system, and geophysical methods as the basic research means, to form the method results of interpreting sedimentary microfacies by classifying seismic waveforms, which is applied to the whole life cycle of exploration and development, solves geological problems of exploration and development, and supports the implementation of improving the quality and efficiency of oil and gas field exploration. The method is simple, has high recognition efficiency, and high interpretation accuracy, and has universal applicability and promotion value in the interpretation of sedimentary microfacies by classifying seismic waveforms.

Claims

1. A method for classifying and interpreting sedimentary microfacies by seismic waveforms, characterized in that Follow these steps: S1, optimization processing of seismic data, taking random noise suppression optimization processing measures to achieve optimization of seismic data resolution and quality; S2, calculation and optimization of three-moment attribute bodies, based on step S1, calculation and selection of seismic three-moment attribute bodies are performed, and the selected attribute body must conform to the knowledge of geological sedimentation, otherwise, steps S1 and S2 are repeated; S3, seismic waveform clustering optimization, optimizes the seismic time window, the number of waveform clusters, and the type of waveform clusterer to achieve seismic waveform clustering; S4, generating a seismic phase map, generating a seismic phase map according to the seismic waveform clustering result, analyzing whether the seismic phase is consistent with the geological knowledge on the generated seismic phase map, and generating a map output if it is consistent, otherwise repeating steps S1 and S2; S5, prepare sedimentary microfacies map, obtain sedimentary microfacies change information based on seismic facies map, combine single well sedimentary microfacies, and use mapping software to prepare sedimentary microfacies map.

2. The method for interpreting sedimentary microfacies by seismic waveform classification according to claim 1, characterized in that In step S1, the seismic data is processed by random noise suppression optimization processing. Random noise suppression of seismic data includes frequency and space. The three-dimensional seismic data is transformed into the frequency-space domain by Fourier transform. The specific operation is as follows: S11, for each frequency f component S(i,j,k), a prediction operator U(l,j,k) is designed using the least squares principle, and its prediction error energy Q(f) is expressed as formula 1: in, is the conjugate of V(i,j,k), L, M, N are the lengths of the longitudinal line, transverse line, and offset of the three-dimensional seismic data space window, respectively, and U(l,m,n) is the three-dimensional space window length prediction operator; S12, by taking partial derivatives of equations 1 and 2, we get the matrix equation: H·U=R Formula 3 Where: H and R are the covariance matrices from 3D seismic data, and U is the 3D prediction operator; S13, for each frequency component, the prediction operator is obtained by solving the matrix equation of formula 3, and then the seismic data is filtered by the prediction operator, and finally converted to the time-space domain by inverse Fourier transform to obtain the result of seismic data denoising.

3. The method for interpreting sedimentary microfacies by seismic waveform classification according to claim 1 or 2, characterized in that In step S2, the earthquake three-moment attribute volume is calculated by obtaining the relationship between earthquake information and geological information over time. The specific operation is as follows: S21, decomposes the seismic wavelet into three items: frequency, phase and amplitude, and its convolution model is: X(t)=s(t)*ε(t) Formula 4 Among them, X(t) is the seismic data, s(t) is the excitation wavelet, and ε(t) is the reflection coefficient; S22, when the reflection coefficient is convolved with the seismic wavelet, the result is described by the three-moment attributes calculated from the seismic data, namely: Instantaneous frequency: Instantaneous phase: θ(t) = arctan[x r (t) / x i (t)] Equation 6 Instantaneous amplitude: Among them, x r (t) is the real part of the seismic data channel, x i (t) is the imaginary seismic data channel, θ(t) is the instantaneous phase, f(t) is the instantaneous frequency, and A(t) is the instantaneous amplitude.

4. The method for interpreting sedimentary microfacies by seismic waveform classification according to any one of claims 1 to 3, characterized in that In step S3, the seismic time window is selected in the range of 50 ms below the bottom interface of the target layer to 50 ms above the top interface.

5. The method for interpreting sedimentary microfacies by seismic waveform classification according to any one of claims 1 to 4, characterized in that In step S3, the number of waveform clusters is 5 to 8.

6. The method for interpreting sedimentary microfacies by classifying seismic waveforms according to any one of claims 1 to 5, characterized in that In step S3, the waveform clusterer type is a supervised waveform classifier or an unsupervised clusterer. The supervised waveform classifier is applicable to the stage with a higher degree of well control, and the unsupervised waveform classifier is applicable to the stage with a lower degree of well control.

7. The method for interpreting sedimentary microfacies by classifying seismic waveforms according to any one of claims 1 to 6, characterized in that In step S4, when generating a seismic phase map, the seismic target layer is divided into different seismic phase zones by utilizing the differences in the characteristics of seismic waveform parameters, and then the sedimentary environment is inferred to generate a seismic phase map.

Images