A method and device for characterizing prior information of a physical property parameter

Abstract

The application provides a physical property parameter prior information feature representation method and device, which comprises the following steps: intercepting the physical property parameter logging curve of a target well in a research area according to preset information; analyzing the characteristic parameters of the physical property parameter logging curve in the form of histogram statistics; constructing a prior distribution probability density function to represent the shape of the histogram; using the characteristic parameter values in the total sample set as the initial values of the component parameters of the prior distribution probability density function; iteratively calculating the component parameters of the prior distribution probability density function, taking the converged component parameters as the final component parameters of the prior distribution probability density function, obtaining the prior distribution probability density function with a deterministic expression, and representing the physical property prior information feature by using the prior distribution probability density function with the deterministic expression. The prior distribution probability density function of the application conforms to the actual situation of the physical property parameter distribution, and provides a guarantee for the accurate prediction of the physical property parameter.

Description

Technical Field

[0001] This invention relates to the field of oil and gas geophysical exploration technology, and in particular to a method and apparatus for characterizing prior information of physical property parameters. Background Technology

[0002] As exploration progresses, oil and gas reservoirs become increasingly complex and concealed, making the discovery of new, potentially valuable reservoirs more difficult, leading to higher exploration costs and increased drilling risks. Oil and gas reservoirs often exhibit strong reflection characteristics in seismic data, such as "bright spots." This seismic response may be caused by high-porosity, high-hydrogen-saturation reservoirs with significant commercial exploitation potential, or by low-porosity, low-hydrogen-saturation reservoirs without commercial exploitation value. Reservoir properties such as porosity and hydrocarbon saturation are crucial for geologists and geophysicists in reservoir evaluation, estimating oil and gas reserves, and determining development well locations.

[0003] In methods for predicting physical property parameters, probabilistic and statistical inversion methods have become the most popular approach due to their combination of the advantages of deterministic and statistical inversion methods. However, these methods rely on prior information features to reduce uncertainty during the inversion process. Therefore, the accurate representation of prior information features is crucial and a prerequisite for the effectiveness of these methods. Conventional representation of prior information features primarily relies on subjective assumptions, assuming the prior characteristics of the physical property parameters to be inverted are Gaussian distributed. A Gaussian distribution is a distribution with a known explicit expression; by simply calculating the mean and variance of the prior information, a specific Gaussian distribution function can be used in the probabilistic and statistical inversion process. Summary of the Invention

[0004] Through extensive experiments, the inventors discovered that the prior characteristics of the actual parameters to be inverted are influenced by factors such as lithology, fluid, pore structure, and interpretation errors. Their distribution characteristics do not perfectly conform to a "single-peak" Gaussian distribution; often, they exhibit "multi-peak" and "long-tail" characteristics. Therefore, directly assuming a Gaussian distribution for the prior characteristics of the physical property parameters to be inverted will introduce significant errors and uncertainties into the subsequent inversion of these parameters. Thus, developing a reasonable solution for characterizing prior information is particularly urgent.

[0005] To address the shortcomings of conventional methods, the present invention aims to provide a method and apparatus for characterizing prior information of physical property parameters, thereby overcoming the problem that existing prior information characterization methods assume that prior information follows a Gaussian distribution, which does not conform to the reality of complex oil and gas reservoirs.

[0006] A method for characterizing prior information of physical property parameters includes the following steps:

[0007] The logging curves of the physical property parameters of the target well in the study area were truncated according to the preset information;

[0008] The physical property parameter logging curves of all wells in the study area are taken as the total sample set. The characteristic parameters of the physical property parameter logging curves are analyzed by histogram statistics on the total sample set. The characteristic parameters include at least the number of peaks, the position of each peak, the ratio of the number of samples corresponding to each peak to the total number of samples, and the dispersion of each peak.

[0009] Construct a prior probability density function to characterize the shape of the histogram;

[0010] The feature parameter values ​​in the total sample set are used as the initial values ​​of the constituent parameters of the prior probability density function;

[0011] The parameters of the prior probability density function are iteratively calculated, and the parameters at convergence are taken as the final parameters of the prior probability density function to obtain a prior probability density function with a deterministic expression. The prior probability density function with a deterministic expression is used to characterize the prior information features of the material properties.

[0012] Furthermore, the logging curves of the physical property parameters of the target well in the study area are truncated according to preset information, including:

[0013] Extract reservoir and non-reservoir segments from the logging curves of the target wells in the study area for all lithological-fluid types.

[0014] Furthermore, the construction of the prior probability density function includes:

[0015] Construct the prior probability density function according to the following equation:

[0016]

[0017] Where x represents the physical property parameter value; M represents the total number of lithological-fluid types in the study area; and k represents one of the lithological-fluid types in the study area. Indicates the weight of type k; This represents the sample mean of physical property parameters belonging to lithology-fluid type k; This represents the sample variance of physical property parameters belonging to lithology-fluid type k; It is pi (π).

[0018] Furthermore, the step of using the feature parameter values ​​from the total sample set as the initial values ​​of the constituent parameters of the prior probability density function includes:

[0019] The number of peaks is used as the initial value of the number of lithofacies-fluid types, the position of each peak is used as the initial value of the mean of the sample values ​​of each lithofacies-fluid physical property parameter, the ratio of the number of samples corresponding to each peak to the total number of samples is used as the initial value of the weight of each lithofacies-fluid type, and the dispersion of each peak is used as the initial value of the variance of the sample values ​​of each lithofacies-fluid physical property parameter.

[0020] Furthermore, the iterative calculation of the component parameters of the prior probability density function, and the use of the component parameters at convergence as the final component parameters of the prior probability density function, includes:

[0021] Obtain the logarithm of the prior probability density function;

[0022] The probability density is calculated using the initial values ​​of the component parameters of the obtained prior probability density function;

[0023] Based on the probability density, the composition parameters of the prior probability density function are recalculated according to the preset iterative formula until the logarithm of the prior probability density function converges. The composition parameters at the point of convergence are then used as the final composition parameters of the prior probability density function.

[0024] Furthermore, the iterative calculation of the component parameters of the prior probability density function, and the use of the component parameters at convergence as the final component parameters of the prior probability density function, includes:

[0025] The logarithm of the prior probability density function is obtained as follows:

[0026]

[0027] Introducing M-dimensional latent variables The probability density is calculated using the initial values ​​of the component parameters of the obtained prior probability density function:

[0028]

[0029] in, ( The value can be 0 or 1. =1 indicates that it is the first Distribution of lithological-fluid types Selected, =0 indicates that no selection is made;

[0030] Based on the probability density, the composition parameters are recalculated according to the following iterative formula:

[0031]

[0032]

[0033]

[0034] in: Indicates the total number of samples, This indicates the number of samples belonging to lithology-fluid type k;

[0035] calculate If convergence is not achieved, the component parameters are recalculated iteratively until convergence is achieved. The component parameters at the point of convergence are then used as the final component parameters of the prior probability density function.

[0036] This invention also provides a device for characterizing prior information of physical property parameters to overcome the problem of inaccurate characterization of prior features of physical properties in existing prior information characterization methods. The device includes a module for extracting logging curves of physical property parameters, a module for analyzing logging curve features of physical property parameters, a module for constructing prior distribution probability density functions, a module for determining initial values ​​of component parameters, and a module for characterizing prior information of physical properties, wherein:

[0037] The physical property parameter logging curve truncation module is used to truncate the physical property parameter logging curves of the target well in the study area according to preset information;

[0038] The physical property parameter logging curve feature analysis module is used to take the physical property parameter logging curves of all wells in the study area as a total sample set, and analyze the characteristic parameters of the physical property parameter logging curves in the form of histogram statistics on the total sample set. The characteristic parameters include at least the number of peaks, the position of each peak, the ratio of the number of samples corresponding to each peak to the total number of samples, and the dispersion of each peak.

[0039] The prior probability density function construction module is used to construct the prior probability density function to characterize the shape of the histogram.

[0040] The initial value determination module for composition parameters is used to use the feature parameter values ​​in the total sample set as the initial values ​​of the composition parameters of the prior distribution probability density function;

[0041] The prior information feature representation module for physical properties is used to iteratively calculate the component parameters of the prior probability density function, and take the component parameters at convergence as the final component parameters of the prior probability density function to obtain a prior probability density function with a deterministic expression. The prior probability density function with a deterministic expression is used to represent the prior information features of physical properties.

[0042] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-mentioned method for characterizing prior information of physical property parameters.

[0043] The present invention also provides a computer-readable storage medium storing a computer program that performs the above-described method for characterizing a priori information of gaseous property parameters.

[0044] This invention also provides a method for predicting oil and gas reservoirs. The prior probability density function with a deterministic expression obtained by the above method is used as a function to represent the prior information of physical property parameters and participates in the inversion of probabilistic and statistical physical property parameters to obtain the inversion results of physical property parameters for predicting oil and gas reservoirs.

[0045] The advantages of this invention compared to existing technologies are as follows: Considering that the prior distribution of actual physical property parameters is not a "single-peak" Gaussian distribution, but rather exhibits "multi-peak" and "long-tail" characteristics, this invention constructs a new prior distribution probability density function with "multi-peak" and "long-tail" characteristics. By determining the constituent parameters of this function through feature statistics and iterative updates, a method and device for characterizing prior information of physical property parameters is developed, resulting in prior feature characterization results that more accurately reflect the actual distribution of physical property parameters, thus ensuring accurate prediction of physical property parameters. The new prior distribution probability density function with "multi-peak" and "long-tail" characteristics constructed in this invention can accurately describe the "non-Gaussian" phenomenon exhibited by actual physical property parameters under the influence of factors such as lithology, fluid, pore structure, and interpretation errors. By using feature statistics as initial values ​​and iteratively determining the constituent parameters of the function, the subjective error problem caused by manually statistically analyzing sample characteristics can be effectively avoided. Attached Figure Description

[0046] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 This is a flowchart of a method for characterizing prior information of physical property parameters in an embodiment of the present invention;

[0048] Figure 2 This is a statistical feature diagram of the gas saturation histogram in an embodiment of the present invention;

[0049] Figure 3 This is a schematic diagram of the distribution function obtained from the gas saturation experiment in an embodiment of the present invention;

[0050] Figure 4 This is a flowchart of step S105 in an embodiment of the present invention;

[0051] Figure 5This is a schematic diagram of a priori information characterization device for physical property parameters in an embodiment of the present invention. Detailed Implementation

[0052] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0053] This embodiment provides a method for characterizing prior information of physical property parameters, such as... Figure 1 As shown, the specific steps include:

[0054] Step S101: Extract the logging curves of the physical property parameters of the target well in the study area according to the preset information.

[0055] If there are multiple wells in the study area, pre-defined information should be extracted from the logging curves of the physical properties of these wells. This pre-defined information extraction can include reservoir and non-reservoir sections of all lithological-fluid types in the study area, such as hydrocarbon-bearing sections, water-bearing sections, and mudstone sections. Because the original logging curves extend from the surface to great depths, it is necessary to extract the curves of the section containing the research target; other sections that are not of interest are considered invalid information.

[0056] Step S102: The physical property parameter logging curves of all wells in the study area are taken as the total sample set. The characteristic parameters of the physical property parameter logging curves are analyzed by histogram statistics on the total sample set. The characteristic parameters include at least the number of peaks, the position of each peak, the ratio of the number of samples corresponding to each peak to the total number of samples, and the dispersion of each peak.

[0057] Specifically, the logging curves of physical property parameters of all wells in the study area are used as the total sample set. The number of peaks, peak positions and dispersion of the samples are automatically recorded, and the characteristics of the physical property parameter samples in the total sample set are analyzed in the form of histogram statistics. Figure 2 The histogram statistics show that the distribution characteristics of the physical property parameters are "multimodal", with two peaks (a common statistical term, meaning that the histogram shows multiple obvious bulges). Figure 2 The "long tail" (a common statistical term) characteristic refers to the length of the sample that extends far from the "peak" and is close to 0 on the horizontal axis of the histogram.

[0058] Step S103: Construct a prior probability density function to characterize the shape of the histogram.

[0059] Construct the prior probability density function according to the following equation:

[0060]

[0061] in,

[0062] x represents the value of the physical property parameter, and its dimension depends on the specific physical property parameter that needs to be characterized.

[0063] M represents the total number of lithological-fluid types in the study area, which is dimensionless;

[0064] k represents one of the lithology-fluid types in the study area, and is dimensionless;

[0065] Indicates the weight of type k;

[0066] This represents the sample mean of physical property parameters belonging to lithology-fluid type k, and its dimensions depend on the specific physical property parameter to be characterized.

[0067] The sample variance of physical property parameters belonging to lithology-fluid type k is dimensionless;

[0068] It is pi, which is dimensionless.

[0069] Step S104: Use the feature parameter values ​​in the total sample set as the initial values ​​of the component parameters of the prior distribution probability density function.

[0070] As can be seen from step S103, the components of the prior probability density function include: the number M of lithofacies-fluid types, and the weight of each lithofacies-fluid type. Mean values ​​of sample values ​​of each lithological-fluid physical property parameter Variance of sample values ​​of each lithological-fluid physical property parameter The statistical results obtained in step S102 are used as the initial values ​​of the parameters of the prior probability density function. Specifically, the number of peaks is used as the initial value of the number of lithofacies-fluid types; the position of each peak is used as the initial value of the mean of the sample values ​​of each lithofacies-fluid physical property parameter; the ratio of the number of samples corresponding to each peak to the total number of samples is used as the initial value of the weight of each lithofacies-fluid type; and the dispersion of each peak is used as the initial value of the variance of the sample values ​​of each lithofacies-fluid physical property parameter.

[0071] Step S105: Iteratively calculate the component parameters of the prior probability density function, and use the component parameters at convergence as the final component parameters of the prior probability density function to obtain a prior probability density function with a deterministic expression. The prior probability density function with a deterministic expression is used to characterize the prior information features of the material properties.

[0072] like Figure 4As shown, this step S105 includes sub-steps S1051-S1053:

[0073] Step S1051: Obtain the logarithm of the prior probability density function.

[0074]

[0075] Step S1052: Calculate the probability density using the initial values ​​of the component parameters of the obtained prior probability density function.

[0076] Introducing M-dimensional latent variables Based on the initial values ​​of the parameters constituting the prior probability density function obtained in step S104, the probability density is calculated:

[0077]

[0078] in, ( The value can be 0 or 1. =1 indicates

[0079] by Distribution of lithological-fluid types Selected; a value of 0 indicates that it is not selected.

[0080] Step S1053: Based on the probability density, recalculate the composition parameters of the prior probability density function according to the preset iterative formula until the logarithm of the prior probability density function converges, and take the composition parameters at the convergence point as the final composition parameters of the prior probability density function.

[0081] according to Recalculate the composition parameters according to the following iterative formula:

[0082]

[0083]

[0084]

[0085] in,

[0086] This indicates the total number of samples.

[0087] This indicates the number of samples belonging to lithology-fluid type k.

[0088] calculate If convergence is not achieved, the component parameters are recalculated iteratively until convergence is achieved. The component parameters at the point of convergence are then used as the final component parameters of the prior probability density function.

[0089] This step uses maximum likelihood theory to fit the unknown parameters in the known expression, and takes the component parameters at convergence as the final component parameters of the prior probability density function. Substituting the final component parameters into the prior probability density function yields a prior probability density function with a deterministic expression.

[0090] Through extensive experiments, the inventors discovered that the prior distribution characteristics of physical property parameters exhibit a "multi-peaked," "long-tailed" shape as shown by the histogram. This invention aims to construct a function that can fit this shape through mathematical theory and experimentation. The shape of this function matches the shape of the histogram, effectively "enclosing" the histogram characteristics, indicating an accurate representation of the histogram's shape. The inventors conducted experiments using gas saturation as an example of physical property parameters, and the resulting distribution function is as follows: Figure 3 As shown by the gray dashed line, this distribution function accurately characterizes the "multimodal" and "long-tailed" features of the physical property parameters. The prior probability density function constructed in this invention differs from any previously known distribution function.

[0091] In practical applications, the prior probability density function with a deterministic expression can be used as a function representing the prior information of physical property parameters, and participate in conventional probabilistic statistical methods for inverting physical property parameters to obtain the inversion results of physical property parameters, which can then serve as the basis for predicting favorable oil and gas reservoirs.

[0092] This invention, based on extensive experimental findings, reveals that the prior distribution of physical property parameters is not a unimodal Gaussian distribution, but rather exhibits multimodal and long-tailed characteristics. Therefore, a new prior distribution probability density function with these multimodal and long-tailed features is constructed. The constituent parameters of this function are determined through feature statistics and iterative updates, resulting in a prior distribution probability density function with a deterministic expression. This prior distribution probability density function conforms to the actual distribution of physical property parameters, providing a guarantee for accurate prediction of these parameters. The new prior distribution probability density function with multimodal and long-tailed characteristics constructed in this invention can accurately describe the non-Gaussian phenomenon exhibited by actual physical property parameters under the influence of factors such as lithology, fluid, pore structure, and interpretation errors. Using feature statistics as initial values ​​and iteratively determining the constituent parameters of the function effectively avoids the subjective errors caused by manually statistically analyzing sample features.

[0093] On the other hand, embodiments of the present invention also provide a device for characterizing prior information of physical property parameters, the structure of which is as follows: Figure 5 As shown, it includes: a module for extracting logging curves of physical property parameters 10, a module for analyzing logging curve features of physical property parameters 20, a module for constructing the prior distribution probability density function 30, a module for determining the initial values ​​of component parameters 40, and a module for characterizing prior information of physical properties 50, wherein:

[0094] The physical property parameter logging curve extraction module 10 is used to extract the physical property parameter logging curves of the target well in the study area according to preset information.

[0095] If there are multiple wells in the study area, pre-defined information should be extracted from the logging curves of the physical properties of these wells. This pre-defined information extraction can include reservoir and non-reservoir sections of all lithological-fluid types in the study area, such as hydrocarbon-bearing sections, water-bearing sections, and mudstone sections. Because the original logging curves extend from the surface to great depths, it is necessary to extract the curves of the section containing the research target; other sections that are not of interest are considered invalid information.

[0096] The physical property parameter logging curve feature analysis module 20 is used to take the physical property parameter logging curves of all wells in the study area as a total sample set, and analyze the characteristic parameters of the physical property parameter logging curves in the form of histogram statistics on the total sample set. The characteristic parameters include at least the number of peaks, the position of each peak, the ratio of the number of samples corresponding to each peak to the total number of samples, and the dispersion of each peak.

[0097] Specifically, the logging curves of physical property parameters of all wells in the study area are used as the total sample set. The number of peaks, peak positions and dispersion of the samples are automatically recorded, and the characteristics of the physical property parameter samples in the total sample set are analyzed in the form of histogram statistics.

[0098] The prior probability density function construction module 30 is used to construct the prior probability density function to characterize the shape of the histogram.

[0099] Construct the prior probability density function according to the following equation:

[0100]

[0101] in,

[0102] x represents the value of the physical property parameter, and its dimension depends on the specific physical property parameter that needs to be characterized.

[0103] M represents the total number of lithological-fluid types in the study area, which is dimensionless;

[0104] k represents one of the lithology-fluid types in the study area, and is dimensionless;

[0105] Indicates the weight of type k;

[0106] This represents the sample mean of physical property parameters belonging to lithology-fluid type k, and its dimensions depend on the specific physical property parameter to be characterized.

[0107] The sample variance of physical property parameters belonging to lithology-fluid type k is dimensionless;

[0108] It is pi, which is dimensionless.

[0109] Through extensive experiments, the inventors discovered that the prior distribution characteristics of physical property parameters exhibit a "multi-peaked," "long-tailed" shape as shown by the histogram. This invention aims to construct a function that can fit this shape through mathematical theory and experimentation. The shape of this function matches the shape of the histogram. Figure 3 The distribution of this function can be seen ( Figure 3 The fact that the gray dashed line "encloses" the histogram features indicates that it accurately represents the shape of the histogram. The prior probability density function constructed in this invention differs from any previously known distribution function.

[0110] The initial value determination module 40 for composition parameters is used to use the feature parameter values ​​in the total sample set as the initial values ​​of the composition parameters of the prior distribution probability density function.

[0111] The statistical results obtained from the well logging curve characteristic analysis module 20 are used as the initial values ​​of the parameters constituting the prior distribution probability density function. Specifically, the number of peaks is used as the initial value for the number of lithofacies-fluid types, and the position of each peak is used as the mean of the sample values ​​of each lithofacies-fluid physical property parameter. The initial values ​​and the positions of each peak are used as the variance of the sample values ​​of each lithological-fluid physical property parameter. Initial value.

[0112] The prior information feature characterization module 50 is used to iteratively calculate the component parameters of the prior probability density function, and take the component parameters at convergence as the final component parameters of the prior probability density function to obtain a prior probability density function with a deterministic expression. The prior probability density function with a deterministic expression is used to characterize the prior information features of the physical properties.

[0113] Specifically, taking the logarithm of the prior probability density function yields:

[0114]

[0115] Introducing M-dimensional latent variables Based on the initial values ​​of the parameters constituting the prior probability density function obtained in step S104, the probability density is calculated:

[0116]

[0117] in, ( The value can be 0 or 1. =1 indicates that it is the first Distribution of lithological-fluid types Selected, =0 indicates no selection. Count the number of selected items. , representing the number of samples belonging to the lithofacies-fluid type k.

[0118] according to Recalculate the composition parameters according to the following iterative formula:

[0119]

[0120]

[0121]

[0122] in,

[0123] This indicates the total number of samples.

[0124] This indicates the number of samples belonging to lithology-fluid type k.

[0125] calculate If convergence is not achieved, the component parameters are recalculated iteratively until convergence is achieved. The component parameters at the point of convergence are then used as the final component parameters of the prior probability density function.

[0126] This step uses maximum likelihood theory to fit the unknown parameters in the known expression, and takes the component parameters at convergence as the final component parameters of the prior probability density function. Substituting the final component parameters into the prior probability density function yields a prior probability density function with a deterministic expression.

[0127] This invention, based on extensive experimental findings, reveals that the prior distribution of physical property parameters is not a unimodal Gaussian distribution, but rather exhibits multimodal and long-tailed characteristics. Therefore, a new prior distribution probability density function with these multimodal and long-tailed features is constructed. The constituent parameters of this function are determined through feature statistics and iterative updates, resulting in a prior distribution probability density function with a deterministic expression. This prior distribution probability density function conforms to the actual distribution of physical property parameters, providing a guarantee for accurate prediction of these parameters. The new prior distribution probability density function with multimodal and long-tailed characteristics constructed in this invention can accurately describe the non-Gaussian phenomenon exhibited by actual physical property parameters under the influence of factors such as lithology, fluid, pore structure, and interpretation errors. Using feature statistics as initial values ​​and iteratively determining the constituent parameters of the function effectively avoids the subjective errors caused by manually statistically analyzing sample features.

[0128] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-mentioned method for characterizing prior information of physical property parameters.

[0129] The present invention also provides a computer-readable storage medium storing a computer program that performs the above-described method for characterizing a priori information of gaseous property parameters.

[0130] This invention also provides a method for predicting oil and gas reservoirs. The prior probability density function with a deterministic expression obtained by the above method is used as a function to represent the prior information of physical property parameters and participates in the inversion of probabilistic and statistical physical property parameters to obtain the inversion results of physical property parameters for predicting oil and gas reservoirs.

[0131] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0132] The above embodiments illustrate that the prior information feature characterization device for physical property parameters of the present invention has flexible scalability and portability.

[0133] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions.

[0134] For ease of description, the above devices are described separately by function as various units. Of course, in implementing this application, the functions of each unit can be implemented in one or more software and / or hardware.

[0135] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory. The computer software product may include several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of this application. The computer software product can be stored in memory, which may include non-permanent memory in computer-readable media, random access memory (RAM), and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media. Computer-readable media includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information may be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0136] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0137] This application can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0138] Although this application has been described by way of examples, those skilled in the art will know that this application has many variations and modifications without departing from the spirit of this application, and it is intended that the appended claims cover such variations and modifications without departing from the spirit of this application.

Claims

1. A method for characterizing prior information of physical property parameters, characterized in that, Includes the following steps: The logging curves of the physical property parameters of the target well in the study area were truncated according to the preset information; The physical property parameter logging curves of all wells in the study area are taken as the total sample set. The characteristic parameters of the physical property parameter logging curves are analyzed by histogram statistics on the total sample set. The characteristic parameters include at least the number of peaks, the position of each peak, the ratio of the number of samples corresponding to each peak to the total number of samples, and the dispersion of each peak. Construct a prior probability density function to characterize the shape of the histogram; The feature parameter values ​​in the total sample set are used as the initial values ​​of the constituent parameters of the prior probability density function; The parameters of the prior probability density function are iteratively calculated, and the parameters at convergence are taken as the final parameters of the prior probability density function to obtain a prior probability density function with a deterministic expression. The prior probability density function with a deterministic expression is used to characterize the prior information features of the physical properties. The prior probability density function is constructed according to the following equation: ； Where x represents the physical property parameter value; M represents the total number of lithological-fluid types in the study area; and k represents one of the lithological-fluid types in the study area. Indicates the weight of type k; This represents the sample mean of physical property parameters belonging to lithology-fluid type k; This represents the sample variance of physical property parameters belonging to lithology-fluid type k; It is pi (π).

2. The method as described in claim 1, characterized in that, The logging curves of the physical property parameters of the target well in the study area are truncated according to preset information, including: Extract reservoir and non-reservoir segments from the logging curves of the target wells in the study area for all lithological-fluid types.

3. The method as described in claim 1, characterized in that, The step of using the feature parameter values ​​from the total sample set as the initial values ​​of the constituent parameters of the prior probability density function includes: The number of peaks is used as the initial value of the number of lithology-fluid types, the position of each peak is used as the initial value of the mean of the sample values ​​of each lithology-fluid physical property parameter, the ratio of the number of samples corresponding to each peak to the total number of samples is used as the initial value of the weight of each lithology-fluid type, and the dispersion of each peak is used as the initial value of the variance of the sample values ​​of each lithology-fluid physical property parameter.

4. The method as described in claim 1, characterized in that, The iterative calculation of the component parameters of the prior probability density function, using the convergent component parameters as the final component parameters of the prior probability density function, includes: Obtain the logarithm of the prior probability density function; The probability density is calculated using the initial values ​​of the component parameters of the obtained prior probability density function; Based on the probability density, the composition parameters of the prior probability density function are recalculated according to the preset iterative formula until the logarithm of the prior probability density function converges. The composition parameters at the point of convergence are then used as the final composition parameters of the prior probability density function.

5. The method as described in claim 1, characterized in that, The iterative calculation of the component parameters of the prior probability density function, using the convergent component parameters as the final component parameters of the prior probability density function, includes: Taking the logarithm of the prior probability density function, we get: ； Introducing M-dimensional latent variables The probability density is calculated using the initial values ​​of the component parameters of the obtained prior probability density function: ； in, The value can be 0 or 1. , =1 indicates that it is the first Distribution of lithological-fluid types Selected, =0 indicates that no selection is made; Based on the probability density, the composition parameters are recalculated according to the following iterative formula: ； ； ； in: Indicates the total number of samples, This indicates the number of samples belonging to lithology-fluid type k; calculate If convergence is not achieved, the component parameters are recalculated iteratively until convergence is achieved. The component parameters at the point of convergence are then used as the final component parameters of the prior probability density function.

6. A device for characterizing prior information of physical property parameters, characterized in that, It includes modules for extracting logging curves of physical property parameters, analyzing logging curve features of physical property parameters, constructing prior distribution probability density functions, determining initial values ​​of component parameters, and characterizing prior information of physical properties. The physical property parameter logging curve truncation module is used to truncate the physical property parameter logging curves of the target well in the study area according to preset information; The physical property parameter logging curve feature analysis module is used to take the physical property parameter logging curves of all wells in the study area as a total sample set, and analyze the characteristic parameters of the physical property parameter logging curves in the form of histogram statistics on the total sample set. The characteristic parameters include at least the number of peaks, the position of each peak, the ratio of the number of samples corresponding to each peak to the total number of samples, and the dispersion of each peak. The prior probability density function construction module is used to construct a prior probability density function to characterize the shape of the histogram. The initial value determination module for the composition parameters is used to use the feature parameter values ​​in the total sample set as the initial values ​​of the composition parameters of the prior distribution probability density function; The prior information feature characterization module for physical properties is used to iteratively calculate the component parameters of the prior probability density function, and uses the component parameters at convergence as the final component parameters of the prior probability density function to obtain a prior probability density function with a deterministic expression. The prior probability density function with a deterministic expression is used to characterize the prior information features of physical properties, wherein: The prior probability density function construction module constructs the prior probability density function according to the following equation: ； Where x represents the physical property parameter value; M represents the total number of lithological-fluid types in the study area; and k represents one of the lithological-fluid types in the study area. Indicates the weight of type k; This represents the sample mean of physical property parameters belonging to lithology-fluid type k; This represents the sample variance of physical property parameters belonging to lithology-fluid type k; It is pi (π).

7. The apparatus as claimed in claim 6, characterized in that, The prior information feature characterization module of physical properties iteratively calculates the component parameters of the prior probability density function, and uses the component parameters at convergence as the final component parameters of the prior probability density function, specifically to obtain the logarithm of the prior probability density function; and calculates the probability density using the initial values ​​of the component parameters of the prior probability density function obtained. Based on the probability density, the composition parameters of the prior probability density function are recalculated according to the preset iterative formula until the logarithm of the prior probability density function converges. The composition parameters at the point of convergence are then used as the final composition parameters of the prior probability density function.

8. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 5.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that performs the method of any one of claims 1 to 5.

10. A method for predicting oil and gas reservoirs, characterized in that, The prior probability density function with a deterministic expression obtained by the method described in any one of claims 1 to 5 is used as a function to represent the prior information of physical property parameters and participates in the inversion of probabilistic and statistical physical property parameters to obtain the inversion results of physical property parameters for predicting oil and gas reservoirs.

Images