Reservoir physical property parameter label synthesis method, apparatus, and training method for prediction model

Abstract

Provided in the present disclosure are a reservoir physical property parameter label synthesis method, an apparatus, and a training method for a prediction model. The reservoir physical property parameter label synthesis method comprises: importing reservoir physical property parameters obtained by means of logging of a target research region into an original physical property parameter distribution characteristic function, so as to obtain a physical property parameter distribution characteristic function corresponding to the target research region; performing random sampling in the physical property parameter distribution characteristic function, so as to obtain a plurality of random synthetic physical property parameters; performing combination according to the plurality of random synthetic physical property parameters, so as to obtain a physical property parameter synthetic curve, and obtaining a reflection coefficient curve; and performing convolution on the reflection coefficient curve and statistical seismic wavelets to obtain a plurality of synthetic pre-stack seismic angle gathers; and, according to the plurality of synthetic pre-stack seismic angle gathers and random synthetic physical property parameters corresponding to each synthetic pre-stack seismic angle gather, determining a plurality of reservoir physical property parameter labels. Since the physical property parameter distribution curve is used to select the reservoir physical property parameters, the difference between the reservoir physical property parameter labels and actual labels is relatively small.

Description

Reservoir physical property parameter label synthesis method, device and prediction model training method

[0001] Related applications

[0002] This application claims priority to the Chinese invention patent application No. 202311518055.7 filed on November 14, 2023, and cites the entire disclosure of the above patent application as part of this application. Technical Field

[0003] The present disclosure relates to the technical field of oil and gas geophysical exploration, and in particular to a reservoir property parameter label synthesis method, device, and prediction model training method. Background Art

[0004] In recent years, the focus of oil and gas exploration has gradually shifted toward lithologic reservoir exploration. Unlike structural reservoirs, these new reservoirs are influenced by structural and reservoir heterogeneity, resulting in complex formation conditions, difficulty in identification, and difficulty in quantitative prediction, posing a high investment risk. In the field of oil and gas geophysical exploration, reservoir physical parameters such as porosity, hydrocarbon saturation, and permeability can quantitatively describe the physical properties of subsurface rocks and are important parameters for lithologic reservoir prediction. Due to the complex formation conditions, a highly nonlinear relationship exists between seismic response characteristics and reservoir physical parameters. Conventional techniques for predicting reservoir physical parameters using seismic data suffer from significant errors and fall short of the requirements for quantitative exploration. The emergence of artificial intelligence technology has made it possible to quantitatively predict the physical parameters of these complex reservoirs.

[0005] In the field of oil and gas geophysical exploration, the integration of artificial intelligence with existing conventional seismic data processing and interpretation techniques has led to intelligent geophysical exploration technology, which can greatly improve the efficiency of seismic data processing and interpretation. Labeled data (including prestack seismic angle gathers and corresponding reservoir physical parameters) is the foundation of artificial intelligence supervised learning networks, and its quantity and quality directly determine the quality of prediction results. Compared with fields such as images and sound, reservoir physical parameters in oil and gas geophysical exploration are more difficult to obtain, and prestack seismic angle gathers are relatively simple. Therefore, the lack of paired labeled data for prestack seismic angle gathers and reservoir physical parameters has seriously restricted the development of intelligent reservoir physical parameter prediction technology. Existing reservoir physical parameter labeling technology generally uses label synthesis methods from the field of image recognition. Label synthesis, based on geological relationships, can synthesize a range of reservoir property values ​​that may exist but are not observed in well logging. Within this range, several reservoir property parameters are randomly selected and the corresponding prestack seismic angle gathers for each reservoir property parameter are calculated using a formula. This yields paired prestack seismic angle gathers and reservoir property parameters, expanding the number of oil and gas (or reservoir) parameter labels in the label set and, to a certain extent, enriching the label information. However, because randomly generated reservoir property parameters are often too theoretical, the synthesized labels are severely distorted and differ significantly from the actual labels, resulting in low accuracy of the trained prediction models. Therefore, developing a generalizable, practical solution for reservoir property parameter label synthesis is particularly urgent.

[0006] Summary of the Invention

[0007] To address the aforementioned issues in the prior art, the present disclosure aims to provide a method and apparatus for synthesizing oil and gas physical property parameter labels, as well as a prediction model training method, to address the prior art issues of severe distortion of synthesized labels, significant differences from actual labels, and low accuracy of trained prediction models. In the present disclosure, parameter labels within the oil and gas physical property parameter label set can be used to describe reservoir properties; namely, the aforementioned oil and gas physical property parameter labels are reservoir physical property parameter labels.

[0008] In order to solve the above technical problems, the specific technical solutions disclosed in this disclosure are as follows:

[0009] In one aspect, the present disclosure provides a method for synthesizing reservoir property parameter labels, comprising:

[0010] Importing the sample set of reservoir physical property parameters obtained from well logging in the target study area into the original physical property parameter distribution characteristic function to obtain the physical property parameter distribution characteristic function corresponding to the target study area;

[0011] Random sampling is performed in the physical property parameter distribution characteristic function to obtain a synthetic sample set of physical property parameters that conforms to the distribution characteristics of the physical property parameters in the study area;

[0012] According to the maximum thickness of the target study area, the seismic sampling rate and the range of lithofacies thickness, the samples in the physical property parameter synthesis sample set are randomly combined to obtain the physical property parameter synthesis curve, and the reflection coefficient curve is obtained based on the physical property parameter synthesis curve;

[0013] Convolve the reflection coefficient curve with the statistical seismic wavelet corresponding to the target study area to obtain several synthetic pre-stack seismic angle gathers;

[0014] The reservoir physical property parameter label data set is determined based on each synthetic pre-stack seismic angle gather and the corresponding physical property parameter synthetic curve.

[0015] As an embodiment of the present disclosure, a sample set of reservoir physical property parameters obtained by well logging in a target study area is imported into an original physical property parameter distribution characteristic function to obtain a physical property parameter distribution characteristic function corresponding to the target study area, further comprising:

[0016] Obtain statistical characteristics of the actual reservoir physical property parameter sample set, including the total number of samples, the total number of lithofacies types, the prior probability of each lithofacies type, and the mean and variance of the physical property parameters belonging to each lithofacies;

[0017] The sample set of reservoir physical property parameters obtained by logging in the target study area is imported into the following original physical property parameter distribution characteristic function to obtain the physical property parameter distribution characteristic function corresponding to the target study area;

[0018] Where x represents the reservoir physical property parameter value, M represents the total number of lithofacies types in the actual sample set of reservoir physical property parameters, the total number of lithofacies types is obtained by comprehensive interpretation of well logging, and the total number of lithofacies types is dimensionless; k represents the kth lithofacies type, dimensionless; α k represents the prior probability of the k-th lithofacies type, which is calculated by counting the proportion of the number of samples belonging to the k-th lithofacies to the total number of samples; μ k represents the mean value of the reservoir physical parameter samples belonging to the kth lithofacies type in the actual sample set of reservoir physical parameter, λ k represents the variance of the reservoir physical property parameter sample of the kth lithofacies type in the actual reservoir physical property parameter sample set; π is the pi, N represents the total number of samples in the actual sample set of physical property parameters; i Represents the i-th sample in the actual sample set of physical property parameters.

[0019] As an embodiment of the present disclosure, obtaining a reflection coefficient curve according to a composite curve of physical parameters further includes:

[0020] Converting the physical property parameter composite curve into the elastic parameter composite curve according to the physical property elasticity conversion equation;

[0021] The elastic parameter synthesis curve is converted into a reflection coefficient curve according to the plane wave reflection equation.

[0022] As an embodiment of the present disclosure, converting a physical property parameter composite curve into an elastic parameter composite curve according to a physical property elasticity conversion equation further includes:

[0023] The physical property parameter synthesis curve is imported into the following physical property elasticity conversion equation to obtain the corresponding elastic parameter curve. The conversion equation is determined based on the rock physical correlation between elastic parameters and physical property parameters under the influence of both lithofacies and pore structure.

[0024] Where x represents the reservoir physical property parameter value; m represents the elastic parameter value; γ j represents the pore structure parameters corresponding to the jth rock in the actual study area, which is obtained from the rock physics experiment in the study area; j (x,γ j ) represents the deterministic multi-porous rock physical model corresponding to the j-th rock; ε j represents the error between the deterministic multi-porous structural rock physics model corresponding to the j-th rock and the actual observation data, j = 1, 2, … M; L(j) is an M × 1 lithofacies constraint vector, where all elements except the j-th one are 0.

[0025] As an embodiment of the present disclosure, after convolving the reflection coefficient curve with the statistical seismic wavelet corresponding to the target study area to obtain a number of synthetic pre-stack seismic angle gathers, the following steps are included:

[0026] Replacing frequency band information of some frequency bands in the synthetic pre-stack seismic angle gather with frequency band data of corresponding frequency bands in the measured pre-stack seismic angle gather to obtain first synthetic pre-stack angle gather data;

[0027] Replacing phase information of some phase segments in the synthetic pre-stack seismic angle gather with phase data of corresponding phase segments in the measured pre-stack seismic angle gather to obtain second synthetic pre-stack angle gather data;

[0028] The first synthetic pre-stack angle gather data and the second synthetic pre-stack angle gather data are synthesized to obtain a synthetic pre-stack seismic angle gather corresponding to the target study area;

[0029] The physical property parameter composite curve and the corresponding synthetic generalized pre-stack seismic angle gather constitute the reservoir physical property parameter label data set.

[0030] As an embodiment of the present disclosure, frequency band information of some frequency bands in a synthetic pre-stack seismic angle gather is replaced with frequency band data of corresponding frequency bands in a measured pre-stack seismic angle gather to obtain first synthetic pre-stack angle gather data, further comprising:

[0031] According to the actual plane distribution of the observation system in the target study area, pre-stack angle gathers of multiple observation points are selected and spectral decomposition is performed to obtain the frequency band data of three types of frequency bands for each observation point to form a spectrum pool;

[0032] Perform spectrum decomposition on synthetic pre-stack seismic angle gathers to obtain frequency band data of three types of frequency bands corresponding to each synthetic pre-stack seismic angle gather;

[0033] The frequency band data of the first frequency band and the frequency band data of the third frequency band of the spectrum data of each synthetic pre-stack angle gather are replaced with the frequency band data of the first frequency band and the frequency band data of the third frequency band randomly selected from the spectrum pool, so as to obtain the first synthetic pre-stack angle gather data of a plurality of frequency domains corresponding to each random synthetic physical property parameter curve;

[0034] Perform spectrum inverse transformation on the frequency domain data of the first synthetic pre-stack angle gather in the frequency domain to obtain the first synthetic pre-stack angle gather data.

[0035] As an embodiment of the present disclosure, replacing phase information of some phase segments in a synthetic pre-stack seismic angle gather with phase data of corresponding phase segments in a measured pre-stack seismic angle gather to obtain second synthetic pre-stack angle gather data further includes:

[0036] According to the actual plane distribution of the observation system in the target study area, the pre-stack angle gathers of multiple observation points are selected and phase decomposition is performed to obtain the phase data of three types of phase segments for each observation point to form a phase pool;

[0037] Perform phase decomposition on synthetic pre-stack seismic angle gathers to obtain phase data of three types of phase segments corresponding to each synthetic pre-stack seismic angle gather;

[0038] The phase data of the first phase segment and the phase data of the third phase segment of each synthetic prestack angle gather are replaced with the phase data of the first phase segment and the phase data of the third phase segment randomly selected from the phase pool to obtain the second synthetic prestack angle gather data of several phase domains corresponding to each random synthetic physical property parameter curve;

[0039] Performing phase inverse transformation on the phase data of the second synthetic pre-stack angle gather in the phase domain to obtain the second synthetic pre-stack angle gather data.

[0040] As an embodiment of the present disclosure, synthesizing the first synthetic pre-stack angle gather data and the second synthetic pre-stack angle gather data to obtain a synthetic pre-stack seismic angle gather corresponding to the target study area further includes:

[0041] Import the first synthetic pre-stack angle gather data and the second synthetic pre-stack angle gather data into the following equation to synthesize the synthetic pre-stack seismic angle gather F corresponding to the target study area. a : F a =F f +bF p

[0042] Among them, F f is the first synthetic pre-stack angle gather data; F p is the second synthetic pre-stack angle gather data; β is the weight coefficient, which is used to adjust the amount of frequency and phase information used.

[0043] On the other hand, the present disclosure also provides a reservoir property parameter label synthesis device, comprising:

[0044] The distribution characteristic calculation unit is used to import the sample set of reservoir physical property parameters obtained by well logging in the target study area into the original physical property parameter distribution characteristic function to obtain the physical property parameter distribution characteristic function corresponding to the target study area;

[0045] The sampling unit is used to perform random sampling in the physical property parameter distribution characteristic function to obtain a synthetic sample set of physical property parameters that conforms to the distribution characteristics of the physical property parameters in the study area;

[0046] The curve synthesis unit is used to randomly combine samples in the physical property parameter synthesis sample set according to the maximum thickness of the target study area, the seismic sampling rate, and the range of lithofacies thickness to obtain a physical property parameter synthesis curve, and to obtain a reflection coefficient curve based on the physical property parameter synthesis curve;

[0047] The seismic trace synthesis unit is used to convolve the reflection coefficient curve with the statistical seismic wavelet corresponding to the target study area to obtain several synthetic pre-stack seismic angle gathers;

[0048] The label determination unit is used to determine a reservoir physical property parameter label data set according to each synthetic pre-stack seismic angle gather and the corresponding physical property parameter synthetic curve.

[0049] On the other hand, this paper also provides a reservoir property parameter prediction model training method, including:

[0050] Using the method of the embodiment of the present disclosure, a reservoir physical property parameter label dataset is obtained;

[0051] Use the reservoir physical property parameter label dataset to train the reservoir physical property parameter prediction model;

[0052] The input of the reservoir physical property parameter prediction model is the pre-stack seismic angle gathers of the exploration area, and the output is the predicted reservoir physical property parameters of the exploration area.

[0053] By adopting the above technical solution, the reservoir physical property parameters obtained by logging in the target study area are imported into the original physical property parameter distribution characteristic function to obtain the physical property parameter distribution characteristic function of the corresponding target study area, so as to determine the physical property parameter distribution curve formed by the reservoir physical property parameters of each well position in the target study area. The physical property parameter distribution curve characterizes the overall reservoir physical property parameter distribution status of the target study area. By randomly sampling the physical property parameter distribution curve to obtain a number of physical property parameter synthetic sample sets, it is possible to obtain random synthetic physical property parameters that meet the reservoir physical property distribution characteristics of the target study area. Compared with the current randomly generated physical property parameters, reservoir physical property parameters that are more in line with the characteristics of the target study area can be obtained. By using the maximum thickness of the target study area, the seismic sampling rate and the range of lithofacies thickness variation, the physical property parameter distribution curve can be obtained. By randomly combining samples in the physical property parameter synthesis sample set to obtain a physical property parameter synthesis curve, and obtaining a reflection coefficient curve based on the physical property parameter synthesis curve, it is possible to obtain a reflection coefficient curve using random synthetic physical property parameters; by convolving the reflection coefficient curve with the statistical seismic wavelet corresponding to the target study area, a number of synthetic pre-stack seismic angle gathers are obtained, and it is possible to obtain the corresponding pre-stack seismic angle gathers using the current random synthetic physical property parameters; by determining the reservoir physical property parameter label data set based on each synthetic pre-stack seismic angle gather and the corresponding physical property parameter synthesis curve, it is possible to obtain a number of synthetic reservoir physical property parameter labels, and since the physical property parameter distribution curve is used to select the reservoir physical property parameters, the difference between the reservoir physical property parameter label and the actual label is small, and the trained prediction model has a high accuracy.

[0054] In order to make the above and other purposes, features and advantages of the present disclosure more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the embodiments of the present disclosure or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0056] FIG1 shows an overall system diagram of a reservoir property parameter label synthesis method according to an embodiment of the present disclosure;

[0057] FIG2 is a schematic diagram showing the steps of a method for synthesizing reservoir physical property parameter labels according to an embodiment of the present disclosure;

[0058] FIG3 shows a schematic diagram of a porosity characterization function according to an embodiment of the present disclosure;

[0059] FIG4 shows a schematic diagram of a composite curve of physical property parameters of mud content according to an embodiment of the present disclosure;

[0060] FIG5 shows a schematic diagram of a composite curve of physical property parameters of porosity according to an embodiment of the present disclosure;

[0061] FIG6 shows a schematic diagram of a synthetic curve of physical property parameters of gas saturation according to an embodiment of the present disclosure;

[0062] FIG7 shows a schematic diagram of a reflection coefficient curve obtained by synthesizing a physical property parameter curve according to an embodiment of the present disclosure;

[0063] FIG8 shows a schematic diagram of adjusting synthetic pre-stack seismic angle gathers according to an embodiment of the present disclosure;

[0064] FIG9 shows a schematic diagram of a reservoir property parameter label synthesis device according to an embodiment of the present disclosure;

[0065] FIG10 shows a schematic diagram of a computer device according to an embodiment of the present disclosure;

[0066] FIG11( a )-FIG11 ( b ) show comparison diagrams of the synthetic pre-stack seismic angle gathers and the actual pre-stack seismic angle gathers according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0067] The following will be combined with the accompanying drawings in the embodiments of the present disclosure to clearly and completely describe the technical solutions in the embodiments of the present disclosure. Obviously, the embodiments described are only part of the embodiments of the present disclosure, not all of the embodiments. Based on the embodiments of the present disclosure, all other embodiments obtained by ordinary technicians in this field without making any creative efforts are within the scope of protection of the present disclosure.

[0068] It should be noted that the terms "first," "second," and the like in the present disclosure, claims, and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or precedence. It should be understood that the numbers used in this manner are interchangeable where appropriate so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having," and any variations thereof, are intended to cover non-exclusive inclusions, for example, a process, method, apparatus, product, or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products, or devices.

[0069] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0070] As shown in FIG1 , the overall system diagram of a reservoir property parameter label synthesis method includes: a parameter memory 101 , an operator 102 , a trainer 103 and a prediction terminal 104 .

[0071] The parameter memory 101 is used to store reservoir physical property parameters obtained by well logging in the target study area. The reservoir physical property parameters obtained by well logging are obtained by downhole exploration using exploration tools, that is, the real reservoir physical property parameters of the target study area.

[0072] The operator 102 is configured to use the reservoir physical property parameters obtained from well logging to fit a physical property parameter distribution curve within the target study area and obtain a corresponding reflection coefficient curve. The operator then convolves the reflection coefficient curve with the statistical seismic wavelet corresponding to the target study area to obtain a number of synthetic pre-stack seismic angle gathers. This is seismic data generated through simulation or calculation is reorganized into multiple data sets according to different incidence angle ranges before being stacked. Each data set contains reflection wave information within a specific angle range. Finally, a number of reservoir physical property parameter labels are determined based on the number of synthetic pre-stack seismic angle gathers and the random synthetic physical property parameters corresponding to each synthetic pre-stack seismic angle gather.

[0073] A trainer 103 is used to train a reservoir property parameter prediction model using a plurality of reservoir property parameter labels;

[0074] The input of the reservoir physical property parameter prediction model is the pre-stack seismic angle gathers of the exploration area, and the output is the reservoir physical property parameters of the exploration area.

[0075] The prediction terminal 104 is used to obtain pre-stack seismic angle gathers of the target exploration area (target study area) and output the corresponding reservoir physical property parameters of the target exploration area (target study area).

[0076] Because the current synthesis method for randomly generating reservoir physical property parameters is too theoretical, the synthesized labels are seriously distorted and differ greatly from the actual labels, and the accuracy of the trained prediction model is low.

[0077] In order to solve the above problems, the embodiment of the present disclosure provides a method for synthesizing reservoir physical property parameter labels, which can solve the problem of large differences between synthesized labels and actual labels. Figure 2 is a step diagram of a method for synthesizing reservoir physical property parameter labels provided by the embodiment of the present disclosure. The present disclosure provides method operation steps such as the embodiment or flow chart, but based on conventional or non-creative labor, more or fewer operation steps may be included. The order of steps listed in the embodiment is only one way of executing the order of many steps and does not represent the only execution order. When the actual system or device product is executed, it can be executed in the order of the method shown in the embodiment or the accompanying drawings or in parallel. Specifically, as shown in Figure 2, the method may include:

[0078] Step 201: importing a sample set of reservoir physical property parameters obtained by well logging in the target study area into the original physical property parameter distribution characteristic function to obtain the physical property parameter distribution characteristic function corresponding to the target study area;

[0079] Step 202: Randomly sample the physical property parameter distribution characteristic function to obtain a physical property parameter synthetic sample set that meets the physical property parameter distribution characteristics of the target study area;

[0080] Step 203: Based on the maximum thickness of the target study area, the seismic sampling rate, and the range of lithofacies thickness variation, randomly combine the samples in the physical property parameter synthesis sample set to obtain a physical property parameter synthesis curve, and obtain a reflection coefficient curve based on the physical property parameter synthesis curve;

[0081] Step 204: convolve the reflection coefficient curve with the statistical seismic wavelet corresponding to the target study area to obtain a number of synthetic pre-stack seismic angle gathers;

[0082] Step 205: Determine a reservoir physical property parameter label data set based on each synthetic pre-stack seismic angle gather and the corresponding physical property parameter synthetic curve.

[0083] By adopting the above technical solution, the reservoir physical property parameters obtained by logging in the target study area are imported into the original physical property parameter distribution characteristic function to obtain the physical property parameter distribution characteristic function of the corresponding target study area, so as to determine the physical property parameter distribution curve formed by the reservoir physical property parameters of each well position in the target study area. The physical property parameter distribution curve characterizes the overall reservoir physical property parameter distribution status of the target study area. By randomly sampling the physical property parameter distribution curve to obtain a number of physical property parameter synthetic sample sets, it is possible to obtain random synthetic physical property parameters that meet the reservoir physical property distribution characteristics of the target study area. Compared with the current randomly generated physical property parameters, reservoir physical property parameters that are more in line with the characteristics of the target study area can be obtained. By using the maximum thickness of the target study area, the seismic sampling rate and the range of lithofacies thickness variation, the physical property parameter distribution curve can be obtained. By randomly combining samples in the physical property parameter synthesis sample set to obtain a physical property parameter synthesis curve, and obtaining a reflection coefficient curve based on the physical property parameter synthesis curve, it is possible to obtain a reflection coefficient curve using random synthetic physical property parameters; by convolving the reflection coefficient curve with the statistical seismic wavelet corresponding to the target study area, a number of synthetic pre-stack seismic angle gathers are obtained, and it is possible to obtain the corresponding pre-stack seismic angle gathers using the current random synthetic physical property parameters; by determining the reservoir physical property parameter label data set based on each synthetic pre-stack seismic angle gather and the corresponding physical property parameter synthesis curve, it is possible to obtain a number of synthetic reservoir physical property parameter labels, and since the physical property parameter distribution curve is used to select the reservoir physical property parameters, the difference between the reservoir physical property parameter label and the actual label is small, and the trained prediction model has a high accuracy.

[0084] As an embodiment of the present disclosure, step 201, importing a sample set of reservoir physical property parameters obtained by well logging in a target study area into an original physical property parameter distribution characteristic function to obtain a physical property parameter distribution characteristic function corresponding to the target study area, further includes:

[0085] Reservoir physical parameters can be parameters such as porosity, oil and gas saturation, and permeability that can quantitatively describe the physical properties of underground rocks.

[0086] Obtain statistical characteristics of the actual reservoir physical property parameter sample set, including the total number of samples, the total number of lithofacies types, the prior probability of each lithofacies type, and the mean and variance of the physical property parameters belonging to each lithofacies;

[0087] The sample set of reservoir physical property parameters obtained by logging in the target study area is imported into the following original physical property parameter distribution characteristic function to obtain the physical property parameter distribution characteristic function corresponding to the target study area;

[0088] Where x represents the reservoir physical property parameter value, M represents the total number of lithofacies types in the actual sample set of reservoir physical property parameters, which is obtained from the comprehensive interpretation of well logging and is dimensionless; k represents the kth lithofacies type, which is dimensionless; α k represents the prior probability of the k-th lithofacies type, which is calculated by counting the proportion of the number of samples belonging to the k-th lithofacies to the total number of samples; μ k represents the mean value of the reservoir physical parameter samples belonging to the kth lithofacies type in the actual sample set of reservoir physical parameter, λ k represents the variance of the reservoir physical parameter sample of the kth lithofacies type in the actual reservoir physical parameter sample set; π is the pi, and N represents the total number of samples x in the actual sample set of physical parameter i Represents the i-th sample in the actual sample set of physical property parameters.

[0089] As shown in Figure 3, the porosity characterization function diagram shows that the vertical direction of the diagram represents the number of sampling points with the same porosity in the target study area, and the horizontal direction represents the porosity. For example, one hundred sets of reservoir physical property parameters are obtained by using exploration equipment to explore the target study area. These one hundred sets of data are respectively substituted into the physical property parameter distribution characteristic function to obtain one hundred points on the distribution characteristic plane (the plane that characterizes the distribution characteristics of the target study area). Then, the one hundred points are used to draw the physical property parameter distribution curve of the target study area. In the distribution curve, it can be seen that the dotted line is the physical property parameter distribution curve, and the columnar data is the reservoir physical property parameter (porosity) obtained by actual logging observation in the target study area. It can be seen that the physical property parameter distribution characteristic function disclosed in the present invention has a good effect in characterizing the target study area, and the trend of the obtained physical property parameter distribution characteristic function is relatively close to the actual reservoir physical property parameter distribution in the target study area.

[0090] Compared with the prior art which uses a target study area to explore and obtain one hundred sets of reservoir physical property parameters, and directly randomly generates reservoir physical property parameters, the present disclosure uses a physical property parameter distribution curve to depict the overall distribution of particular physical property parameters at different locations in the target study area, and performs random sampling in the obtained physical property parameter distribution curve to obtain a number of randomly synthesized physical property parameters. Therefore, the randomly synthesized physical property parameters finally obtained by the present disclosure are more consistent with the actual situation of the target study area, and thus the finally obtained reservoir physical property parameter labels can be more consistent with the actual reservoir physical property parameter labels, so that the trained prediction model can have a higher accuracy rate.

[0091] As an embodiment of the present disclosure, step 202 is to perform random sampling in the physical property parameter distribution characteristic function to obtain a physical property parameter synthetic sample set that meets the physical property parameter distribution characteristics of the study area.

[0092] In this step, the Markov chain Monte Carlo random sampling method is used to perform random sampling in the physical property parameter distribution characteristic function obtained by fitting, so as to obtain a large number of random synthetic physical property parameters that conform to the physical property parameter distribution characteristics, and form a physical property parameter synthetic sample set that conforms to the physical property parameter distribution characteristics of the study area.

[0093] As an embodiment of the present disclosure, based on the maximum thickness of the target study area, the seismic sampling rate, and the range of lithofacies thickness variation, samples in the physical property parameter synthesis sample set are randomly combined to obtain a physical property parameter synthesis curve, including:

[0094] According to the maximum time thickness of the overall target layer in the target study area, the seismic sampling rate in the target study area and the thickness variation range of each lithofacies, the physical parameters are randomly combined from shallow to deep to obtain a large number of physical parameter synthetic curves with the same length.

[0095] As shown in Figure 4, the schematic diagram of the synthetic curve of physical property parameters of mud content, the schematic diagram of the synthetic curve of physical property parameters of porosity, and the schematic diagram of the synthetic curve of physical property parameters of gas saturation, it can be seen that the shapes of the synthetic physical property parameter curves vary, which can better simulate the values ​​of different physical property parameters.

[0096] As shown in FIG7 , a schematic diagram of a reflection coefficient curve obtained by synthesizing a physical property parameter curve is provided. As an embodiment of the present disclosure, obtaining a reflection coefficient curve according to the physical property parameter synthesizing curve further includes:

[0097] Step 701: Convert the physical property parameter composite curve into the elastic parameter composite curve according to the physical property elasticity conversion equation;

[0098] According to the correlation between lithofacies and pore structure, the physical property elasticity conversion equation is established for the study area with n types of lithofacies.

[0099] The physical property parameter synthesis curve is imported into the following physical property elasticity conversion equation to obtain the corresponding elastic parameter curve. The conversion equation is determined based on the rock physical correlation between elastic parameters and physical property parameters under the influence of both lithofacies and pore structure.

[0100] Where x represents the reservoir physical property parameter value; m represents the elastic parameter value; γ j represents the pore structure parameters corresponding to the jth rock in the actual study area, which is obtained from the rock physics experiment in the study area; j (x,γ j ) represents the deterministic multi-porous rock physical model corresponding to the j-th rock; ε j represents the error between the deterministic multi-porous structure rock physics model corresponding to the j-th rock and the actual observation data, j = 1, 2, ... M; L(j) is an M × 1 lithofacies constraint vector, in which all elements except the j-th value are 0

[0101] The elastic parameter composite curve of the target study area is drawn based on several elastic parameter values.

[0102] Step 702: Convert the elastic parameter synthesis curve into a reflection coefficient curve according to the plane wave reflection equation.

[0103] The elastic parameter synthesis curve is converted into a reflection coefficient curve using the Zoeppritz plane wave reflection equation.

[0104] As an embodiment of the present disclosure, the reflection coefficient curve is convolved with the statistical seismic wavelet corresponding to the target study area to obtain several synthetic pre-stack seismic angle gathers, specifically including:

[0105] Based on the seismic wave convolution theory, the incident angle range of the pre-stack angle gather is set, and the reflection coefficient curve is convolved with the statistical seismic wavelet of the near, medium and far angle partial stack gathers in the target study area to obtain a large number of corresponding synthetic pre-stack seismic angle gathers.

[0106] Then, based on the multiple synthetic prestack seismic angle gathers and the randomly synthesized physical property parameters corresponding to each synthetic prestack seismic angle gather, multiple reservoir physical property parameter labels are determined. At this point, reservoir physical property parameter labels are obtained, where the reservoir physical property parameters included in the reservoir physical property parameter labels are randomly generated based on the physical property parameter distribution curve, and the corresponding synthetic prestack seismic angle gathers are generated according to a formula. Therefore, compared to the prior art method of directly randomly generating reservoir physical property parameters and using randomly generated reservoir physical property parameters to generate synthetic prestack seismic angle gathers, the reservoir physical property parameter labels generated by the present disclosure are more consistent with the actual conditions of the target study area.

[0107] It should be noted that, in order to simplify the description, the pre-stack angle gather in this disclosure refers to the pre-stack seismic angle gather.

[0108] As shown in FIG8 , which is a schematic diagram of adjusting synthetic pre-stack seismic angle gathers, as one embodiment of the present disclosure, after convolving the reflection coefficient curve with the statistical seismic wavelet corresponding to the target study area to obtain several synthetic pre-stack seismic angle gathers, the following steps are included:

[0109] Step 801: Replace the frequency band information of some frequency bands in the synthetic pre-stack seismic angle gather with the frequency band data of the corresponding frequency bands of the measured pre-stack seismic angle gather to obtain first synthetic pre-stack angle gather data.

[0110] First, the pre-stack angle gathers in the time domain are converted to the frequency domain.

[0111] According to the actual plane distribution of the observation system in the target study area, pre-stack angle gathers of multiple observation points are selected and spectral decomposition is performed to obtain frequency band data of three types of frequency bands for each observation point to form a spectrum pool; in the present disclosure, the frequency bands can be divided into three categories according to the length of the frequency bands. For example, if the total length of the frequency band of each observation point is 50 Hz, then 1-15 Hz can be set as the first type of frequency band data, 16-35 Hz can be set as the second type of frequency band data, and 36-50 Hz can be set as the third type of frequency band data.

[0112] First, the synthetic pre-stack seismic angle gathers in the time domain are converted to the frequency domain.

[0113] The synthetic pre-stack seismic angle gathers are spectrally decomposed to obtain frequency band data of three types of frequency bands corresponding to each synthetic pre-stack seismic angle gather; since the present disclosure convolves the reflection coefficient curve with the statistical seismic wavelet corresponding to the target study area to obtain several synthetic pre-stack seismic angle gathers, the total frequency band length of the synthetic pre-stack seismic angle gathers is consistent with the total frequency band length of the pre-stack seismic angle gathers of each observation point in the actual observation system. Therefore, the synthetic pre-stack seismic angle gathers can also be divided into three categories at equal intervals according to the length of the frequency bands. For example, if the total frequency band length of each synthetic pre-stack seismic angle gather is 50Hz, then 1-15Hz can be set as the first type of frequency band data, 16-35Hz can be set as the second type of frequency band data, and 36-50Hz can be set as the third type of frequency band data.

[0114] The frequency band data of the first frequency band and the frequency band data of the third frequency band of the spectrum data of each synthetic pre-stack angle gather are replaced with the frequency band data of the first frequency band and the frequency band data of the third frequency band randomly selected from the spectrum pool, so as to obtain the first synthetic pre-stack angle gather data of a plurality of frequency domains corresponding to each random synthetic physical property parameter curve;

[0115] In the present disclosure, since the spectrum pool contains several first-category frequency band data and third-category frequency band data obtained from actual observations of the same target study area, such as first-category frequency band data A, first-category frequency band data B, first-category frequency band data C; third-category frequency band data E, third-category frequency band data F, and third-category frequency band data G, if the spectrum data of a synthetic pre-stack angle gather includes first-category frequency band data 1, second-category frequency band data 2, and third-category frequency band data 3. Since low-frequency data in a frequency band data segment usually corresponds to background data in the data, and high-frequency data usually corresponds to noise data in the data, it is easier to reflect actual characteristics. The present invention uses frequency band data obtained from different actual observations to replace the spectral data of the synthetic pre-stack angle gather. For example, the first-class frequency band data A and the third-class frequency band data E are used to replace the first-class frequency band data 1 and the third-class frequency band data 3, and the spectral data of the synthetic pre-stack angle gather finally obtained includes the first-class frequency band data A, the second-class frequency band data 2, and the third-class frequency band data E; similarly, the first-class frequency band data A and the third-class frequency band data F can also be used to replace the first-class frequency band data 1 and the third-class frequency band data 3, and the spectral data of the synthetic pre-stack angle gather finally obtained includes the first-class frequency band data A, the second-class frequency band data 2, and the third-class frequency band data F. By analogy, a synthetic pre-stack angle gather in which a plurality of first-class frequency band data and third-class frequency band data are replaced with the actually observed frequency band data can be obtained.

[0116] Perform spectrum inverse transformation on the frequency domain data of the first synthetic pre-stack angle gather in the frequency domain to obtain the first synthetic pre-stack angle gather data.

[0117] Step 802: Replace the phase information of some phase segments in the synthetic pre-stack seismic angle gather with the phase data of the corresponding phase segments in the measured pre-stack seismic angle gather to obtain second synthetic pre-stack angle gather data.

[0118] First, the prestack angle gathers in the time domain are converted to the phase domain.

[0119] According to the actual plane distribution of the observation system in the target study area, the pre-stack angle gathers of multiple observation points are selected and phase decomposition is performed to obtain the phase data of three types of phase segments for each observation point to form a phase pool;

[0120] In the present disclosure, the phase segments can be divided into three categories according to the length of the phase. For example, if the total length of the phase segment of each observation point is 60°, then 1-20° can be set as the first category of phase segment data, 21-40° can be set as the second category of phase segment data, and 41-60° can be set as the third category of phase segment data.

[0121] First, the synthetic pre-stack seismic angle gathers in the time domain are converted to the phase domain.

[0122] Perform phase decomposition on synthetic pre-stack seismic angle gathers to obtain phase data of three types of phase segments corresponding to each synthetic pre-stack seismic angle gather;

[0123] Since the present invention convolves the reflection coefficient curve with the statistical seismic wavelet corresponding to the target study area to obtain several synthetic pre-stack seismic angle gathers, the total length of the phase segment of the synthetic pre-stack seismic angle gathers is consistent with the total length of the phase segment of the pre-stack seismic angle gathers of each observation point in the actual observation system. Therefore, the synthetic pre-stack seismic angle gathers can also be divided into three categories according to the length of the phase segment. For example, if the total length of the phase segment of each synthetic pre-stack seismic angle gather is 60°, then 1-20° can be set as the first category of phase segment data, 21-40° can be set as the second category of phase segment data, and 41-60° can be set as the third category of phase segment data.

[0124] The phase data of the first phase segment and the phase data of the third phase segment of each synthetic prestack angle gather are replaced with the phase data of the first phase segment and the phase data of the third phase segment randomly selected from the phase pool to obtain the second synthetic prestack angle gather data of several phase domains corresponding to each random synthetic physical property parameter curve;

[0125] In the present disclosure, since the phase pool contains phase data of the first type of phase segment and phase data of the third type of phase segment obtained from actual observations of the same target study area, such as first type of phase segment data A, first type of phase segment data B, first type of phase segment data C; third type of phase segment data E, third type of phase segment data F, and third type of phase segment data G; if the phase data of a synthetic pre-stack angle gather includes first type of phase segment data 1, second type of phase segment data 2, and third type of phase segment data 3. Since low phase segment data in a phase segment data usually corresponds to background data in the data, and high phase segment data usually corresponds to noise data in the data, it is easier to reflect the characteristics in practice. The present invention uses different phase data obtained from actual observations to replace the phase data of the synthetic pre-stack angle gather. For example, the first-class phase segment data A and the third-class phase segment data E are used to replace the first-class phase segment data 1 and the third-class phase segment data 3, and the phase data of the synthetic pre-stack angle gather finally obtained includes the first-class phase segment data A, the second-class phase segment data 2, and the third-class phase segment data E; similarly, the first-class phase segment data A and the third-class phase segment data F can also be used to replace the first-class phase segment data 1 and the third-class phase segment data 3, and the phase data of the synthetic pre-stack angle gather finally obtained includes the first-class phase segment data A, the second-class phase segment data 2, and the third-class phase segment data F. By analogy, a synthetic pre-stack angle gather in which a plurality of first phase segment data and third phase segment data are replaced by the actually observed phase segment data can be obtained.

[0126] Performing phase inverse transformation on the phase data of the second synthetic pre-stack angle gather in the phase domain to obtain the second synthetic pre-stack angle gather data.

[0127] Step 803: synthesize the first synthetic pre-stack angle gather data and the second synthetic pre-stack angle gather data to obtain a synthetic pre-stack seismic angle gather corresponding to the target study area;

[0128] Import the first synthetic pre-stack angle gather data and the second synthetic pre-stack angle gather data into the following equation to synthesize the synthetic pre-stack seismic angle gather F corresponding to the target study area. a : F a =F f +bF p

[0129] Among them, F f is the first synthetic pre-stack angle gather data; F p is the second synthetic pre-stack angle gather data; β is the weight coefficient, which is used to adjust the amount of frequency and phase information used.

[0130] As shown in the comparison diagrams of the synthetic pre-stack seismic angle gathers and the actual pre-stack seismic angle gathers in Figures 11(a) and 11(b), it can be seen that the synthetic pre-stack seismic angle gathers synthesized in the present invention are highly similar to the actual seismic angle gathers. Therefore, the reservoir physical property parameter labels disclosed in the present invention are more consistent with the target study area.

[0131] Step 804: The physical property parameter composite curve and the corresponding synthetic generalized pre-stack seismic angle gather constitute a reservoir physical property parameter label dataset.

[0132] The physical property parameter composite curve and the corresponding synthetic generalized pre-stack seismic angle gather constitute the reservoir physical property parameter label data set; specifically, using a pair of calculated reservoir physical property parameters and synthetic pre-stack seismic angle gathers, multiple types of generalized synthetic pre-stack seismic angle gathers can be generated, that is, one-to-many data are generated through one-to-one data, which increases the number of reservoir physical property parameter labels. Moreover, since the generalized synthetic pre-stack seismic angle gathers replace the pre-stack seismic angle gathers obtained by actual observation, the final reservoir physical property parameter labels are closer to the actual situation of the target study area.

[0133] For different study areas, the lithofacies type, mean and variance in the characteristic characterization function, seismic wavelet, spectrum and phase pool of the actual pre-stack seismic angle gather are updated to obtain generalized label data suitable for the study area.

[0134] FIG9 is a schematic diagram of a reservoir property parameter label synthesis device, comprising:

[0135] The distribution characteristic calculation unit 901 is used to import the sample set of reservoir physical property parameters obtained by well logging in the target study area into the original physical property parameter distribution characteristic function to obtain the physical property parameter distribution characteristic function corresponding to the target study area;

[0136] The sampling unit 902 is used to perform random sampling in the physical property parameter distribution characteristic function to obtain a physical property parameter synthetic sample set that meets the physical property parameter distribution characteristics of the target study area;

[0137] The curve synthesis unit 903 is used to randomly combine samples in the physical property parameter synthesis sample set according to the maximum thickness of the target study area, the seismic sampling rate, and the range of lithofacies thickness to obtain a physical property parameter synthesis curve, and to obtain a reflection coefficient curve based on the physical property parameter synthesis curve;

[0138] The seismic trace synthesis unit 904 is used to convolve the reflection coefficient curve with the statistical seismic wavelet corresponding to the target study area to obtain a number of synthetic pre-stack seismic angle gathers;

[0139] The label determination unit 905 is configured to determine a reservoir physical property parameter label data set based on each synthetic pre-stack seismic angle gather and the corresponding physical property parameter synthetic curve.

[0140] By adopting the above technical solution, the reservoir physical property parameters obtained by logging in the target study area are imported into the original physical property parameter distribution characteristic function to obtain the physical property parameter distribution characteristic function of the corresponding target study area, so as to determine the physical property parameter distribution curve formed by the reservoir physical property parameters of each well position in the target study area. The physical property parameter distribution curve characterizes the overall reservoir physical property parameter distribution status of the target study area. By randomly sampling the physical property parameter distribution curve to obtain a number of physical property parameter synthetic sample sets, it is possible to obtain random synthetic physical property parameters that meet the reservoir physical property distribution characteristics of the target study area. Compared with the current randomly generated physical property parameters, reservoir physical property parameters that are more in line with the characteristics of the target study area can be obtained. By using the maximum thickness of the target study area, the seismic sampling rate and the range of lithofacies thickness variation, the physical property parameter distribution curve can be obtained. By randomly combining samples in the physical property parameter synthesis sample set to obtain a physical property parameter synthesis curve, and obtaining a reflection coefficient curve based on the physical property parameter synthesis curve, it is possible to obtain a reflection coefficient curve using random synthetic physical property parameters; by convolving the reflection coefficient curve with the statistical seismic wavelet corresponding to the target study area, a number of synthetic pre-stack seismic angle gathers are obtained, and it is possible to obtain the corresponding pre-stack seismic angle gathers using the current random synthetic physical property parameters; by determining the reservoir physical property parameter label data set based on each synthetic pre-stack seismic angle gather and the corresponding physical property parameter synthesis curve, it is possible to obtain a number of synthetic reservoir physical property parameter labels, and since the physical property parameter distribution curve is used to select the reservoir physical property parameters, the difference between the reservoir physical property parameter label and the actual label is small, and the trained prediction model has a high accuracy.

[0141] The present disclosure also provides a reservoir property parameter prediction model training method, comprising:

[0142] The reservoir physical property parameter label dataset is obtained by using the method;

[0143] Use the reservoir physical property parameter label dataset to train the reservoir physical property parameter prediction model;

[0144] The input of the reservoir physical property parameter prediction model is the pre-stack seismic angle gathers of the exploration area, and the output is the predicted reservoir physical property parameters of the exploration area.

[0145] As shown in FIG10 , a computer device 1002 is provided according to an embodiment of the present disclosure. The computer device 1002 may include one or more processors 1004, such as one or more central processing units (CPUs), each of which may implement one or more hardware threads. The computer device 1002 may also include any memory 1006 for storing any type of information, such as code, settings, data, and the like. For example, and without limitation, the memory 1006 may include any one or more combinations of the following: any type of RAM, any type of ROM, a flash memory device, a hard disk, an optical disk, and the like. More generally, any memory may use any technology to store information. Furthermore, any memory may provide volatile or non-volatile retention of information. Furthermore, any memory may represent a fixed or removable component of the computer device 1002. In one embodiment, when the processor 1004 executes associated instructions stored in any memory or combination of memories, the computer device 1002 may perform any operation of the associated instructions. The computer device 1002 also includes one or more drive mechanisms 1008, such as a hard disk drive mechanism, an optical disk drive mechanism, and the like, for interacting with any memory.

[0146] The computer device 1002 may also include an input / output module 1010 (I / O) for receiving various inputs (via input devices 1012) and for providing various outputs (via output devices 1014). A specific output mechanism may include a presentation device 1016 and an associated graphical user interface (GUI) 1018. In other embodiments, the input / output module 1010 (I / O), input devices 1012, and output devices 1014 may not be included, and the computer device 1002 may simply be a computer device in a network. The computer device 1002 may also include one or more network interfaces 1020 for exchanging data with other devices via one or more communication links 1022. One or more communication buses 1024 couple the components described above together.

[0147] The communication link 1022 may be implemented in any manner, for example, via a local area network, a wide area network (e.g., the Internet), a point-to-point connection, etc., or any combination thereof. The communication link 1022 may include any combination of hardwired links, wireless links, routers, gateway functions, name servers, etc., governed by any protocol or combination of protocols.

[0148] Corresponding to the method in FIG. 2 to FIG. 8 , an embodiment of the present disclosure further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above method are executed.

[0149] The embodiment of the present disclosure also provides a computer-readable instruction, wherein when a processor executes the instruction, the program therein causes the processor to execute the method shown in Figures 2 to 8.

[0150] It should be understood that in the various embodiments of the present disclosure, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present disclosure.

[0151] It should also be understood that in the embodiments of this disclosure, the term "and / or" merely describes the relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A exists alone, A and B exist simultaneously, or B exists alone. Furthermore, the character " / " in this disclosure generally indicates that the associated objects are in an "or" relationship.

[0152] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in this disclosure can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the composition and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this disclosure.

[0153] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0154] In the several embodiments provided in the present disclosure, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, or can be an electrical, mechanical or other form of connection.

[0155] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, i.e., they may be located in one place or distributed across multiple network units. Some or all of these units may be selected based on actual needs to achieve the objectives of the embodiments of the present disclosure.

[0156] In addition, the functional units in the various embodiments of the present disclosure may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0157] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present disclosure is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the various embodiments of the present disclosure. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0158] Specific embodiments are used in the present disclosure to illustrate the principles and implementation methods of the present disclosure. The description of the above embodiments is only used to help understand the method and core ideas of the present disclosure. At the same time, for those skilled in the art, according to the ideas of the present disclosure, there may be changes in the specific implementation methods and application scopes. In summary, the content of the present disclosure should not be understood as a limitation of the present disclosure.

Claims

1. A reservoir property parameter label synthesis method, characterized in that: include: Importing a sample set of reservoir physical property parameters obtained by logging in the target study area into the original physical property parameter distribution characteristic function to obtain the physical property parameter distribution characteristic function corresponding to the target study area; Random sampling is performed in the physical property parameter distribution characteristic function to obtain a physical property parameter synthetic sample set that meets the physical property parameter distribution characteristics of the study area; According to the maximum thickness of the target study area, the seismic sampling rate and the range of lithofacies thickness variation, the samples in the physical property parameter synthesis sample set are randomly combined to obtain a physical property parameter synthesis curve, and a reflection coefficient curve is obtained according to the physical property parameter synthesis curve; Convolving the reflection coefficient curve with the statistical seismic wavelet corresponding to the target study area to obtain a number of synthetic pre-stack seismic angle gathers; According to each synthetic pre-stack seismic angle gather and the corresponding physical property parameter synthetic curve, a reservoir physical property parameter label data set is determined.

2. The reservoir property parameter label synthesis method according to claim 1, characterized in that: The step of importing the sample set of reservoir physical property parameters obtained by logging in the target study area into the original physical property parameter distribution characteristic function to obtain the physical property parameter distribution characteristic function corresponding to the target study area further includes: The statistical characteristics of the actual reservoir physical property parameter sample set are obtained by statistics, including the total number of samples, the total number of lithofacies types, the prior probability of each lithofacies type, and the mean and variance of the physical property parameters belonging to each lithofacies; Importing a sample set of reservoir physical property parameters obtained by logging in the target study area into the following original physical property parameter distribution characteristic function, to obtain a physical property parameter distribution characteristic function corresponding to the target study area; Wherein, x represents the value of reservoir physical property parameter, M represents the total number of lithofacies types in the actual sample set of reservoir physical property parameter, the total number of lithofacies types is obtained by comprehensive interpretation of well logging, and the total number of lithofacies types is dimensionless; k represents the kth lithofacies type, dimensionless; ɑ k represents the prior probability of the k-th lithofacies type, which is calculated by counting the proportion of the number of samples belonging to the k-th lithofacies to the total number of samples; μ k represents the mean value of the reservoir physical parameter samples belonging to the kth lithofacies type in the actual sample set of reservoir physical parameter, λ k Represents the reservoir physical property parameter sample of the kth lithofacies type in the actual reservoir physical property parameter sample set Variance; π is the ratio of circumference to circumference, and N represents the total number of samples in the actual sample set of physical property parameters x i Represents the i-th sample in the actual sample set of physical property parameters.

3. The reservoir property parameter label synthesis method according to claim 1, characterized in that: The step of obtaining a reflection coefficient curve according to the physical property parameter synthesis curve further comprises: Converting the physical property parameter composite curve into an elastic parameter composite curve according to a physical property elasticity conversion equation; The elastic parameter synthesis curve is converted into a reflection coefficient curve according to the plane wave reflection equation.

4. The reservoir property parameter label synthesis method according to claim 3, characterized in that: The step of converting the physical property parameter composite curve into the elastic parameter composite curve according to the physical property elasticity conversion equation further comprises: The physical property parameter synthesis curve is introduced into the following physical property elastic conversion equation to obtain the corresponding elastic parameter curve, wherein the conversion equation is determined based on the rock physics correlation between the elastic parameters and the physical property parameters under the influence of the two factors of lithofacies and pore structure; Where x represents the reservoir physical property parameter value; m represents the elastic parameter value; γ j represents the pore structure parameters corresponding to the jth rock in the actual study area, obtained from the rock physics experiment in the study area; j (x,γ j ) represents the deterministic multi-porous structure rock physics model corresponding to the j-th rock; ε j It represents the error between the deterministic multi-porous structure rock physics model corresponding to the j-th rock and the actual observation data, j = 1, 2, ... M; L(j) is an M × 1 lithofacies constraint vector, in which all elements except the j-th value are 0.

5. The reservoir property parameter label synthesis method according to claim 1, characterized in that: After convolving the reflection coefficient curve with the statistical seismic wavelet corresponding to the target study area to obtain a plurality of synthetic pre-stack seismic angle gathers, the method further includes: The frequency band information of some frequency bands in the synthetic pre-stack seismic angle gather is replaced with the frequency band data of the corresponding frequency bands of the measured pre-stack seismic angle gather to obtain the first synthetic pre-stack angle gather data; The phase information of some phase segments in the synthetic pre-stack seismic angle gather is replaced with the phase data of the corresponding phase segments in the measured pre-stack seismic angle gather to obtain the second synthetic pre-stack angle gather data; The first synthetic pre-stack angle gather data and the second synthetic pre-stack angle gather data are synthesized to obtain a synthetic pre-stack seismic angle gather corresponding to the target study area; The physical property parameter composite curve and the corresponding synthetic generalized pre-stack seismic angle gather constitute the reservoir physical property parameter label data set.

6. The reservoir property parameter label synthesis method according to claim 5, characterized in that: The step of replacing the frequency band information of some frequency bands in the synthetic pre-stack seismic angle gather with the frequency band data of the corresponding frequency bands of the measured pre-stack seismic angle gather to obtain the first synthetic pre-stack angle gather data further comprises: According to the actual plane distribution of the observation system in the target study area, prestack angle gathers of multiple observation points are selected and spectral decomposition is performed to obtain frequency band data of three types of frequency bands of each observation point to form a spectrum pool; Perform spectrum decomposition on synthetic pre-stack seismic angle gathers to obtain frequency band data of three frequency bands corresponding to each synthetic pre-stack seismic angle gather; The frequency band data of the first frequency band and the frequency band data of the third frequency band of the spectrum data of each synthetic pre-stack angle gather are replaced with the frequency band data of the first frequency band and the frequency band data of the third frequency band randomly selected from the spectrum pool to obtain a plurality of first synthetic pre-stack angle gather data in frequency domains corresponding to each random synthetic physical property parameter curve; Performing spectrum inverse transformation on the frequency domain data of the first synthetic pre-stack angle gather in the frequency domain to obtain the first synthetic pre-stack angle gather data.

7. The reservoir property parameter label synthesis method according to claim 5, characterized in that: The step of replacing the phase information of some phase segments in the synthetic pre-stack seismic angle gather with the phase data of the corresponding phase segments of the measured pre-stack seismic angle gather to obtain the second synthetic pre-stack angle gather data further comprises: According to the actual plane distribution of the observation system in the target study area, the prestack angle gathers of multiple observation points are selected and phase decomposition is performed to obtain the phase data of three types of phase segments of each observation point to form a phase pool; Perform phase decomposition on synthetic pre-stack seismic angle gathers to obtain phase data of three types of phase segments corresponding to each synthetic pre-stack seismic angle gather; The phase data of the first type of phase segment and the phase data of the third type of phase segment of each phase segment data of the synthetic pre-stack angle gather are compared with the phase data of the first type of phase segment and the phase data of the third type of phase segment randomly selected from the phase pool. The phase data of the segment are replaced to obtain the second synthetic pre-stack angle gather data of several phase domains corresponding to each random synthetic physical property parameter curve; Performing a phase inverse transformation on the second synthetic pre-stack angle gather phase data in the phase domain to obtain the second synthetic pre-stack angle gather data.

8. The reservoir property parameter label synthesis method according to claim 5, characterized in that: The synthesizing the first synthetic pre-stack angle gather data and the second synthetic pre-stack angle gather data to obtain a synthetic pre-stack seismic angle gather corresponding to the target study area further comprises: The first synthetic pre-stack angle gather data and the second synthetic pre-stack angle gather data are introduced into the following equation to synthesize the synthetic pre-stack seismic angle gather F corresponding to the target study area. a : F a =F f +βF p Among them, F f is the first synthetic pre-stack angle gather data; F p is the second synthetic pre-stack angle gather data; β is the weight coefficient, which is used to adjust the amount of frequency and phase information adopted.

9. A reservoir property parameter label synthesis device, characterized in that: include: A distribution characteristic calculation unit is used to import the sample set of reservoir physical property parameters obtained by logging in the target study area into the original physical property parameter distribution characteristic function to obtain the physical property parameter distribution characteristic function corresponding to the target study area; A sampling unit is used to perform random sampling in the physical property parameter distribution characteristic function to obtain a physical property parameter synthetic sample set that meets the physical property parameter distribution characteristics of the study area; A curve synthesis unit, for randomly combining samples in a physical property parameter synthesis sample set to obtain a physical property parameter synthesis curve according to the maximum thickness of the target study area, the seismic sampling rate and the range of lithofacies thickness variation, and obtaining a reflection coefficient curve according to the physical property parameter synthesis curve; A seismic trace synthesis unit, used for convolving the reflection coefficient curve with the statistical seismic wavelet corresponding to the target study area to obtain a number of synthetic pre-stack seismic angle gathers; The label determination unit is used to determine a reservoir physical property parameter label data set according to each synthetic pre-stack seismic angle gather and a corresponding physical property parameter synthetic curve.

10. A reservoir physical property parameter prediction model training method, characterized in that: include: Using the method according to any one of claims 1 to 8, obtaining a reservoir physical property parameter label data set; Use reservoir physical property parameter label data set to train reservoir physical property parameter prediction model; The input of the reservoir physical property parameter prediction model is the pre-stack seismic angle gathers of the exploration area, and the output is the reservoir physical property parameters predicted in the exploration area.

Images