Permeability prediction method, device and equipment

Abstract

The invention provides a permeability prediction method, device and equipment, and the method comprises the steps: obtaining a plurality of training sample data, each training sample data comprises the longitudinal wave speed, the transverse wave speed, the density, the longitudinal wave attenuation factor and the transverse wave attenuation factor of a reservoir, and the permeability corresponding to the reservoir; obtaining a permeability prediction model by adopting a machine learning method according to the multiple training sample data; acquiring longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor and transverse wave attenuation factor of the target reservoir; and inputting the longitudinal wave velocity, the transverse wave velocity, the density, the longitudinal wave attenuation factor and the transverse wave attenuation factor of the target reservoir into the permeability prediction model to obtain the permeability of the target reservoir. In the permeability prediction process, not only is the relationship between the elastic parameter and the permeability considered, but also the relationship between the attenuation information and the permeability is fully considered, the information dimension of the data is increased, and the permeability prediction precision is improved.

Description

Technical Field

[0001] The embodiments of the present invention relate to the technical field of oil exploration, and particularly to a method, device and equipment for predicting permeability. Background Art

[0002] Permeability is one of the important petrophysical parameters, and accurate permeability evaluation is of great significance for reservoir evaluation and productivity prediction. At present, the prediction of permeability is mainly calculated through well logging curves. Specifically, the permeability curve is obtained based on elastic parameters, and then the permeability of the reservoir is predicted using the permeability curve. However, since the change in permeability contributes very little to the change in elastic parameters, even less than the contribution of saturation, it is very difficult to establish a non-linear relationship between conventional elastic parameters and permeability. This results in a very large error in the permeability predicted based on elastic parameters, making it difficult to meet the actual application requirements. Summary of the Invention

[0003] The embodiments of the present invention provide a method, device and equipment for predicting permeability to solve the problem of large error in the permeability predicted by the existing methods.

[0004] In a first aspect, the embodiments of the present invention provide a method for predicting permeability, including:

[0005] Obtain a plurality of training sample data, each training sample data including the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor and transverse wave attenuation factor of the reservoir, and the permeability corresponding to the reservoir;

[0006] Use the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor and transverse wave attenuation factor in each training sample data as the input of the model, and use the corresponding permeability in each training sample data as the expected output of the model, and obtain a permeability prediction model by using the machine learning method;

[0007] Obtain the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor and transverse wave attenuation factor of the target reservoir;

[0008] Input the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor and transverse wave attenuation factor of the target reservoir into the permeability prediction model to obtain the permeability of the target reservoir.

[0009] In one embodiment, obtaining a plurality of training sample data includes:

[0010] Obtain well logging data, the well logging data including longitudinal wave velocity, transverse wave velocity, density, porosity, shale content and saturation curves;

[0011] Perform petrophysical modeling based on the well logging data to obtain a petrophysical model;

[0012] Determine the longitudinal wave attenuation factor and the shear wave attenuation factor according to the rock physics model.

[0013] In one embodiment, taking the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor, and shear wave attenuation factor in each training sample data as the input of the model, and taking the corresponding permeability in each training sample data as the expected output of the model, a permeability prediction model is obtained by using machine learning methods, including:

[0014] Establish a permeability prediction model using the support vector machine method to reflect the non-linear relationship between the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor, and shear wave attenuation factor and permeability.

[0015] In one embodiment, obtaining the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor, and shear wave attenuation factor of the target reservoir includes:

[0016] Obtain the longitudinal wave velocity, shear wave velocity, and density of the target reservoir through pre-stack three-parameter inversion;

[0017] Obtain the longitudinal wave attenuation factor and shear wave attenuation factor of the target reservoir through pre-stack amplitude-versus-frequency inversion.

[0018] In a second aspect, an embodiment of the present invention provides a permeability prediction device, including:

[0019] A collection module for obtaining a plurality of training sample data, each training sample data including the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor, and shear wave attenuation factor of the reservoir, and the permeability corresponding to the reservoir;

[0020] A training module for taking the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor, and shear wave attenuation factor in each training sample data as the input of the model, and taking the corresponding permeability in each training sample data as the expected output of the model, and obtaining a permeability prediction model by using machine learning methods;

[0021] An acquisition module for obtaining the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor, and shear wave attenuation factor of the target reservoir;

[0022] A prediction module for inputting the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor, and shear wave attenuation factor of the target reservoir into the permeability prediction model to obtain the permeability of the target reservoir.

[0023] In one embodiment, the collection module is used to obtain a plurality of training sample data, including for:

[0024] Obtain logging data, which includes longitudinal wave velocity, shear wave velocity, density, porosity, shale content, and saturation curves;

[0025] Perform rock physics modeling based on logging data to obtain a rock physics model;

[0026] Determine the longitudinal wave attenuation factor and the transverse wave attenuation factor according to the rock physics model.

[0027] In one embodiment, a training module is configured to use the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor in each training sample data as the input of the model, and use the corresponding permeability in each training sample data as the expected output of the model, and obtain a permeability prediction model by using machine learning methods, including:

[0028] Establish a permeability prediction model by using the support vector machine method to reflect the non-linear relationship between the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor and the permeability.

[0029] In one embodiment, an acquisition module is configured to acquire the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor of a target reservoir, including:

[0030] Acquire the longitudinal wave velocity, transverse wave velocity, and density of the target reservoir through prestack three-parameter inversion;

[0031] Acquire the longitudinal wave attenuation factor and the transverse wave attenuation factor of the target reservoir through prestack amplitude-versus-frequency inversion.

[0032] In a third aspect, an embodiment of the present invention provides an electronic device, including:

[0033] At least one processor and a memory;

[0034] The memory stores computer-executable instructions;

[0035] At least one processor executes the computer-executable instructions stored in the memory, so that at least one processor executes the permeability prediction method according to any one of the first aspects.

[0036] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, in which computer-executable instructions are stored, and when the computer-executable instructions are executed by a processor, they are used to implement the permeability prediction method according to any one of the first aspects.

[0037] The permeability prediction method, device and equipment provided by the embodiments of the present invention obtain a plurality of training sample data, where each training sample data includes the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor and transverse wave attenuation factor of the reservoir, and the permeability corresponding to the reservoir; use the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor and transverse wave attenuation factor in each training sample data as the input of the model, and use the corresponding permeability in each training sample data as the expected output of the model, and obtain a permeability prediction model by using the method of machine learning; obtain the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor and transverse wave attenuation factor of the target reservoir; input the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor and transverse wave attenuation factor of the target reservoir into the permeability prediction model to obtain the permeability of the target reservoir. In the process of permeability prediction, not only the relationship between elastic parameters and permeability is considered, but also the relationship between attenuation information and permeability is fully considered, increasing the information dimension of the data and improving the accuracy of permeability prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] The accompanying drawings herein are incorporated into the specification and constitute a part of this specification, showing embodiments consistent with the present invention and used together with the specification to explain the principles of the present invention.

[0039] Figure 1 It is a flowchart of the permeability prediction method provided by an embodiment of the present invention;

[0040] Figure 2 It is a flowchart of the permeability prediction method provided by another embodiment of the present invention;

[0041] Figure 3 It is a flowchart of the permeability prediction method provided by another embodiment of the present invention;

[0042] Figure 4 It is a schematic structural diagram of the permeability prediction device provided by an embodiment of the present invention;

[0043] Figure 5 It is a schematic structural diagram of the electronic device provided by an embodiment of the present invention.

[0044] Through the above-mentioned accompanying drawings, specific embodiments of the present invention have been shown, and there will be more detailed descriptions hereinafter. These drawings and textual descriptions are not intended to limit the scope of the inventive concept in any way, but to illustrate the concept of the present invention to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0045] The present invention will be further described in detail below in conjunction with the accompanying drawings through specific embodiments. Similar elements in different embodiments are denoted by related similar element numbers. In the following embodiments, many details are described to enable a better understanding of the present application. However, those skilled in the art can easily recognize that some of the features can be omitted in different situations, or can be replaced by other elements, materials, or methods. In some cases, some operations related to the present application are not shown or described in the specification to avoid overwhelming the core part of the present application with excessive description. For those skilled in the art, it is not necessary to describe these related operations in detail, and they can fully understand the related operations based on the description in the specification and the general technical knowledge in the art.

[0046] In addition, the features, operations, or characteristics described in the specification can be combined in any suitable manner to form various embodiments. At the same time, the steps or actions in the method description can also be reordered or adjusted in a manner obvious to those skilled in the art. Therefore, the various sequences in the specification and the drawings are only for clearly describing a certain embodiment and do not mean a necessary sequence unless it is stated that a certain sequence must be followed.

[0047] The serial numbers assigned to the components herein, such as "first", "second", etc., are only used to distinguish the described objects and do not have any sequential or technical meaning. The "connection" and "coupling" mentioned in the present application, unless otherwise specified, both include direct and indirect connection (coupling).

[0048] To solve the problem of large prediction errors in existing permeability prediction methods, this application establishes the relationship between elastic parameters and viscoelastic parameters and permeability, and realizes the prediction of permeability based on the pre-stack amplitude versus incidence angle and frequency (AVF) inversion, improving the accuracy of permeability prediction. Conventional pre-stack inversion can obtain conventional elastic parameters (longitudinal wave velocity, transverse wave velocity, and density). If one wants to predict permeability, the non-linear relationship between longitudinal wave velocity, transverse wave velocity, and density and permeability needs to be obtained. However, since the influence of permeability itself on longitudinal wave velocity, transverse wave velocity, and density is very small, it is difficult to obtain this non-linear relationship. This application makes full use of the obvious corresponding relationship between viscoelastic parameters (longitudinal wave attenuation factor Qp and transverse wave attenuation factor Qs) and permeability. By introducing the longitudinal wave attenuation factor Qp and transverse wave attenuation factor Qs of viscoelastic parameters, together with the conventional elastic parameters of longitudinal wave velocity, transverse wave velocity, and density, a non-linear relationship between these five parameters and permeability is established, and this non-linear relationship is then used for the permeability prediction of the target reservoir, finally obtaining more accurate permeability data. Further, considering the lack of sample quantity of logging data, in order to improve the generalization ability and obtain more accurate permeability, when establishing the non-linear relationship between these five parameters and permeability, this application adopts the method based on Support Vector Machine (SVM).

[0049] Support Vector Machine (SVM) is a pattern recognition method based on statistical learning theory. It is based on the VC dimension theory and the principle of structural risk minimization of statistical learning theory. It seeks the best compromise between the complexity of the model (i.e., the learning accuracy for specific training samples) and the learning ability (i.e., the ability to correctly identify any sample without error) according to limited sample information, in order to obtain the best generalization ability. The support vector machine regression algorithm maps the data to a high-dimensional feature space through a kernel function and performs linear regression in this space. The linear regression in the high-dimensional feature space corresponds to the non-linear regression in the low-dimensional input space.

[0050] Let the training sample set be {(x i ,y i ), i = 1, 2, …, l}, where x i ∈R N is the N-dimensional measured model sample input value, y i ∈R N is the sample output value, and l is the number of samples. For the above training sample set, through non-linear mapping Map the sample data x in the training set to a high-dimensional linear feature space, and construct a linear regression estimation function in this linear space whose dimension may be infinite. Assume that the estimation function is in the following form:

[0051]

[0052] where w∈R NH , b∈R, is a nonlinear mapping that maps the input space to a high-dimensional feature space; b is the bias. The purpose of the solution is to find the parameter w T , b, so that for input x outside the sample, there is

[0053]

[0054] The optimization problem corresponding to formula (1) is The constraints are:

[0055]

[0056] Introduce slack variables ε,ε * , in order to ensure that the above equation has a solution, (2) is transformed into:

[0057]

[0058] The constraints are:

[0059]

[0060] ε≥0,ε i * ≥0,i=1,2,…,l (3)

[0061] Here C>0 is the penalty coefficient. The larger C is, the greater the penalty for data points that exceed the error ε. ε is an insensitive loss function, which is in the form of:

[0062]

[0063] Obviously, equation (4) is a constrained quadratic programming. Next, we use the Lagrange multiplier method to solve this quadratic programming with linear inequality constraints, namely:

[0064]

[0065] where a i , a i * ≥0, i=1,2,…,l, is the Lagrange multiplier.

[0066] Introducing the kernel function K(x i ,x j)Instead of non - linear mapping The kernel function is any symmetric function that satisfies the Mercer condition. From this, the dual optimization problem of equation (5) can be obtained:

[0067]

[0068] The constraint conditions are:

[0069]

[0070] By solving equation (6), the final estimation function can be obtained as:

[0071]

[0072] When using support vector machines for linear regression, when the number of samples is large, the learning speed is relatively slow, while when the number of samples is scarce, the generalization ability is very good. That is to say, support vector machines have strong small - sample learning ability. Therefore, in this application, the method based on support vector machines is used to obtain the non - linear relationship regarding permeability, which helps to further improve the accuracy of permeability prediction. The following will use specific embodiments to elaborate on the method provided in this application in detail.

[0073] Embodiment 1

[0074] Figure 1 is the flowchart of the permeability prediction method provided by an embodiment of the present invention. As Figure 1 shown, the permeability prediction method provided in this embodiment may include:

[0075] S101. Obtain a plurality of training sample data, where each training sample data includes the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor of the reservoir, as well as the corresponding permeability of the reservoir.

[0076] To realize the training of the permeability prediction model, it is first necessary to collect training sample data. In this embodiment, the training sample data can be obtained through multiple rock samples in the research area. It should be emphasized that the sample data obtained in this embodiment not only includes elastic parameters (longitudinal wave velocity, transverse wave velocity, and density), but also includes visco - elastic parameters (i.e., attenuation information, longitudinal wave attenuation factor, and transverse wave attenuation factor). Seismic amplitude is dependent on permeability. By considering the double - porosity mesoscopic flow model, it is found that at any given frequency, there is a linear relationship between attenuation information and permeability. That is to say, attenuation information is very sensitive to changes in permeability. Therefore, permeability can be predicted through attenuation information.

[0077] First, well logging data of the samples can be collected, including the longitudinal wave velocity, transverse wave velocity, density, porosity, shale content, and saturation curve. Then, petrophysical modeling of the viscoelastic medium can be completed based on the well logging data. Finally, the longitudinal wave attenuation factor Qp curve and the transverse wave attenuation factor Qs curve can be calculated based on the established petrophysical model. When modeling the petrophysical model of seismic wave attenuation, the compression modulus under low-frequency and high-frequency conditions is considered separately. When calculating the compression modulus under low-frequency conditions, the Backus average of the dry rock compression modulus of each part and the algebraic average of the porosity are substituted into the fluid substitution equation for calculation. The compression modulus under high-frequency conditions can be obtained by taking the Backus average of the compression modulus of each part. The difference between the high-frequency and low-frequency moduli reflects the intrinsic attenuation caused by local fluid flow induced by elastic inhomogeneity in the non-reservoir section of the porous rock. Based on this, the Qp (longitudinal wave attenuation factor) and Qs (transverse wave attenuation factor) curves are calculated. That is to say, in an alternative implementation, obtaining a plurality of training sample data may specifically include: obtaining well logging data, where the well logging data includes the longitudinal wave velocity, transverse wave velocity, density, porosity, shale content, and saturation curve; performing petrophysical modeling based on the well logging data to obtain a petrophysical model; and determining the longitudinal wave attenuation factor and the transverse wave attenuation factor based on the petrophysical model.

[0078] S102: Using the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor in each training sample data as the input of the model, and using the corresponding permeability in each training sample data as the expected output of the model, a permeability prediction model is obtained by using machine learning methods.

[0079] After obtaining a plurality of training sample data, these training sample data can be used to train the permeability prediction model. Specifically, during training, the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor in each training sample data are used as the input of the model, the corresponding permeability in each training sample data is used as the expected output of the model, and the difference between the actual output and the expected output of the permeability prediction model is used as the loss function. Training is completed when the value of the loss function is less than a preset threshold. That is to say, the actual output of the model is made to approximate the expected output as much as possible. In this embodiment, the specific implementation manner of the permeability prediction model is not limited.

[0080] S103: Obtain the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor of the target reservoir.

[0081] After training a permeability prediction model based on training sample data, the model can be used to predict the permeability of the target reservoir. To predict the permeability of the target reservoir, it is first necessary to obtain the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor of the target reservoir. In this embodiment, there is no limitation on how to obtain the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor of the target reservoir. For example, it can be obtained by measurement, calculation, consulting materials, etc.

[0082] S104. Input the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor of the target reservoir into the permeability prediction model to obtain the permeability of the target reservoir.

[0083] After obtaining the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor of the target reservoir, input these parameters into the permeability prediction model trained in step S102, and the permeability of the target reservoir can be determined through the permeability prediction model.

[0084] The permeability prediction method, device, and equipment provided in this embodiment obtain multiple training sample data, where each training sample data includes the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor of the reservoir, as well as the permeability corresponding to the reservoir; use the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor in each training sample data as the input of the model, and use the corresponding permeability in each training sample data as the expected output of the model, and obtain the permeability prediction model by using the machine learning method; obtain the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor of the target reservoir; input the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor of the target reservoir into the permeability prediction model to obtain the permeability of the target reservoir. In the process of permeability prediction, not only the relationship between elastic parameters and permeability is considered, but also the relationship between attenuation information and permeability is fully considered, increasing the information dimension of the data and improving the accuracy of permeability prediction.

[0085] Embodiment 2

[0086] Considering that the number of samples that can be obtained in the research area is limited, on the basis of the above embodiment, in order to further improve the prediction accuracy of the permeability prediction model, the support vector machine method with strong small sample learning ability is used to construct the permeability prediction model in this embodiment. The support vector machine can be used as a regression method to fit each parameter. Using the support vector machine method, the permeability and parameters such as longitudinal and transverse wave velocities, density, Qp, and Qs are mapped to a high-dimensional space through a kernel function, which is convenient for finding the linear relationship between them, and ensures the permeability prediction accuracy in the case of limited sample quantity.

[0087] Figure 2 The flowchart of the permeability prediction method provided by another embodiment of the present invention. As Figure 2 shown, the permeability prediction method provided in this embodiment may specifically include:

[0088] S201. Obtain a plurality of training sample data, each training sample data including the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor and shear wave attenuation factor of the reservoir, and the permeability corresponding to the reservoir.

[0089] S202. Establish a permeability prediction model by using the support vector machine method to reflect the non-linear relationship between the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor, shear wave attenuation factor and permeability.

[0090] S203. Obtain the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor and shear wave attenuation factor of the target reservoir.

[0091] S204. Input the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor and shear wave attenuation factor of the target reservoir into the permeability prediction model to obtain the permeability of the target reservoir.

[0092] Based on the above embodiment, the permeability prediction method provided in this embodiment further establishes a permeability prediction model by using the support vector machine method to reflect the non-linear relationship between the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor, shear wave attenuation factor and permeability, which can further improve the permeability prediction accuracy in the case of limited sample quantity.

[0093] Embodiment III

[0094] Based on any of the above embodiments, how to obtain the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor and shear wave attenuation factor of the target reservoir will be further elaborated below. Specifically, the longitudinal wave velocity, shear wave velocity and density of the target reservoir can be obtained through prestack three-parameter inversion, and here the information of the change of amplitude with offset is mainly used; the longitudinal wave attenuation factor and shear wave attenuation factor of the target reservoir can be obtained through AVF inversion, and here the information of the change of amplitude with frequency is mainly used. Figure 3 The flowchart of the permeability prediction method provided by another embodiment of the present invention. As Figure 3 shown, the permeability prediction method provided in this embodiment may specifically include:

[0095] S301. Obtain a plurality of training sample data, each training sample data including the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor and shear wave attenuation factor of the reservoir, and the permeability corresponding to the reservoir.

[0096] S302. Using the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor, and shear wave attenuation factor in each training sample data as the input of the model, and using the corresponding permeability in each training sample data as the expected output of the model, a permeability prediction model is obtained by using machine learning methods.

[0097] S303. Obtain the longitudinal wave velocity, shear wave velocity, and density of the target reservoir through prestack three-parameter inversion.

[0098] S304. Obtain the longitudinal wave attenuation factor and shear wave attenuation factor of the target reservoir through prestack amplitude-versus-frequency inversion.

[0099] S305. Input the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor, and shear wave attenuation factor of the target reservoir into the permeability prediction model to obtain the permeability of the target reservoir.

[0100] Based on any of the above embodiments, the permeability prediction method provided in this embodiment further obtains the longitudinal wave velocity, shear wave velocity, and density of the target reservoir through prestack three-parameter inversion, making full use of the variation information of amplitude with offset; obtains the longitudinal wave attenuation factor and shear wave attenuation factor of the target reservoir through prestack amplitude-versus-frequency inversion, making full use of the variation information of amplitude with frequency, increasing the information dimension of the data, and making the accuracy of the permeability prediction result higher.

[0101] In summary, the permeability prediction method provided in this application aims to solve the problem of large prediction errors of permeability in existing methods. In permeability prediction, it not only considers the relationship between elastic parameters (longitudinal wave velocity, shear wave velocity, density) and permeability, but also simultaneously considers the relationship between attenuation information (longitudinal wave attenuation factor and shear wave attenuation factor) and permeability, increasing the information dimension of the data and making the accuracy of the permeability prediction result higher. This is because, generally speaking, prestack elastic parameter inversion uses the variation information of amplitude with offset and does not use the information in the frequency domain, while this application considers frequency information, which is equivalent to adding another dimension of information, that is, frequency information, to the previous variation information of amplitude with offset. Therefore, it is an improvement in dimension and increases the information dimension of the data. Further, considering that the number of samples that can be obtained in the study area is limited, a support vector machine method with strong small-sample learning ability is used to construct the permeability prediction model, ensuring the accuracy of permeability prediction even when the number of samples is limited.

[0102] Embodiment 4

[0103] Figure 4 It is a schematic structural diagram of a permeability prediction device provided in an embodiment of the present invention. As Figure 4 shown, the permeability prediction device 40 provided in this embodiment may include:

[0104] A collection module 401, configured to obtain a plurality of training sample data, where each training sample data includes the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor, and shear wave attenuation factor of a reservoir, and the permeability corresponding to the reservoir;

[0105] A training module 402, configured to use the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor, and shear wave attenuation factor in each training sample data as the input of the model, and use the corresponding permeability in each training sample data as the expected output of the model, and obtain a permeability prediction model by using a machine learning method;

[0106] An acquisition module 403, configured to obtain the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor, and shear wave attenuation factor of a target reservoir;

[0107] A prediction module 404, configured to input the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor, and shear wave attenuation factor of the target reservoir into the permeability prediction model to obtain the permeability of the target reservoir.

[0108] The device in this embodiment can be used to execute Figure 1 the technical solution of the method embodiment shown, and its implementation principle and technical effect are similar, and will not be elaborated here.

[0109] In an optional implementation manner, the specific operations for the collection module 401 to obtain a plurality of training sample data may include:

[0110] Obtain logging data, where the logging data includes longitudinal wave velocity, shear wave velocity, density, porosity, shale content, and saturation curves;

[0111] Perform rock physics modeling based on the logging data to obtain a rock physics model;

[0112] Determine the longitudinal wave attenuation factor and shear wave attenuation factor according to the rock physics model.

[0113] In an optional implementation manner, the specific operations for the training module 402 to use the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor, and shear wave attenuation factor in each training sample data as the input of the model, and use the corresponding permeability in each training sample data as the expected output of the model, and obtain a permeability prediction model by using a machine learning method may include:

[0114] Establish a permeability prediction model using the support vector machine method to reflect the non-linear relationship between longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor, and shear wave attenuation factor and permeability.

[0115] In an optional implementation manner, the specific operations for the acquisition module 403 to obtain the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor, and shear wave attenuation factor of a target reservoir may include:

[0116] Obtain the P-wave velocity, S-wave velocity and density of the target reservoir through prestack three-parameter inversion;

[0117] Obtain the P-wave attenuation factor and S-wave attenuation factor of the target reservoir through prestack amplitude-versus-frequency inversion.

[0118] Example 5

[0119] The embodiment of the present invention also provides an electronic device. Please refer to Figure 5 As shown, the embodiment of the present invention only takes Figure 5 as an example for illustration, and does not mean that the present invention is limited thereto. Figure 5 It is a schematic structural diagram of an electronic device provided by an embodiment of the present invention. As Figure 5 shown, the electronic device 50 provided in this embodiment may include: a memory 501, a processor 502, and a bus 503. Among them, the bus 503 is used to realize the connection between various components.

[0120] A computer program is stored in the memory 501, and when the computer program is executed by the processor 502, the technical solutions of any of the above method embodiments can be realized.

[0121] Among them, the memory 501 and the processor 502 are directly or indirectly electrically connected to realize data transmission or interaction. For example, these components can be electrically connected to each other through one or more communication buses or signal lines, such as being connected through the bus 503. A computer program for realizing the permeability prediction method is stored in the memory 501, including at least one software function module that can be stored in the memory 501 in the form of software or firmware. The processor 502 executes various functional applications and data processing by running the software programs and modules stored in the memory 501.

[0122] The memory 501 can be, but is not limited to, a Random Access Memory (RAM), a Read Only Memory (ROM), a Programmable Read-Only Memory (PROM), an Erasable Programmable Read-Only Memory (EPROM), an Electric Erasable Programmable Read-Only Memory (EEPROM), etc. Among them, the memory 501 is used to store programs, and after receiving an execution instruction, the processor 502 executes the programs. Further, the software programs and modules in the memory 501 may also include an operating system, which may include various software components and / or drivers for managing system tasks (such as memory management, storage device control, power management, etc.), and may communicate with various hardware or software components to provide a running environment for other software components.

[0123] The processor 502 can be an integrated circuit chip with signal processing capabilities. The above-mentioned processor 502 can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc. It can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of the present invention. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. It can be understood that Figure 5 the structure of is only schematic, and it may also include more or fewer components than Figure 5 shown in, or have a different configuration from Figure 5 shown in. Figure 5 Each component shown in can be implemented by hardware and / or software.

[0124] The embodiments of the present invention also provide a computer-readable storage medium, on which a computer program is stored, and the computer program is executed by a processor to implement the technical solutions of any one of the above method embodiments.

[0125] Each embodiment in this disclosure is described in a progressive manner. For the same or similar parts among the embodiments, reference can be made to each other. Each embodiment focuses on the differences from other embodiments.

[0126] The protection scope of the present disclosure is not limited to the above embodiments. Obviously, those skilled in the art can make various modifications and deformations to the present disclosure without departing from the scope and spirit of the present disclosure. If these modifications and deformations fall within the scope of the claims of the present disclosure and their equivalent technologies, the intention of the present disclosure also includes these modifications and deformations.

Claims

1. A method for predicting permeability, characterized in that, it includes: Obtain a plurality of training sample data, each training sample data including the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor of the reservoir, and the permeability corresponding to the reservoir; Using the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor in each training sample data as the input of the model, and the corresponding permeability in each training sample data as the expected output of the model, obtain a permeability prediction model by using machine learning methods; Obtain the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor of the target reservoir; Input the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor of the target reservoir into the permeability prediction model to obtain the permeability of the target reservoir.

2. The method according to claim 1, characterized in that, the obtaining of a plurality of training sample data includes: Obtain well logging data, the well logging data including longitudinal wave velocity, transverse wave velocity, density, porosity, shale content, and saturation curves; Conduct petrophysical modeling based on the well logging data to obtain a petrophysical model; Determine the longitudinal wave attenuation factor and transverse wave attenuation factor according to the petrophysical model.

3. The method according to claim 1, characterized in that, the using of the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor in each training sample data as the input of the model, and the corresponding permeability in each training sample data as the expected output of the model, and obtaining a permeability prediction model by using machine learning methods includes: Establish a permeability prediction model by using the support vector machine method to reflect the non-linear relationship between the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor and permeability.

4. The method according to any one of claims 1-3, characterized in that, the obtaining of the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor of the target reservoir includes: Obtain the longitudinal wave velocity, transverse wave velocity, and density of the target reservoir through prestack three-parameter inversion; Obtain the longitudinal wave attenuation factor and transverse wave attenuation factor of the target reservoir through prestack amplitude-versus-frequency inversion.

5. A permeability prediction device, characterized in that, it includes: A collection module for obtaining a plurality of training sample data, each training sample data including the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor of the reservoir, and the permeability corresponding to the reservoir; A training module for using the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor in each training sample data as the input of the model, and the corresponding permeability in each training sample data as the expected output of the model, and obtaining a permeability prediction model by using machine learning methods; An obtaining module for obtaining the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor of the target reservoir; A prediction module for inputting the longitudinal wave velocity, transverse wave velocity, density, longitudinal wave attenuation factor, and transverse wave attenuation factor of the target reservoir into the permeability prediction model to obtain the permeability of the target reservoir.

6. The permeability prediction device according to claim 5, wherein, the collection module is configured to obtain a plurality of training sample data, including: obtaining well logging data, where the well logging data includes longitudinal wave velocity, shear wave velocity, density, porosity, shale content, and saturation curve; performing rock physics modeling based on the well logging data to obtain a rock physics model; determining a longitudinal wave attenuation factor and a shear wave attenuation factor according to the rock physics model.

7. The permeability prediction device according to claim 5, wherein, the training module is configured to use the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor, and shear wave attenuation factor in each training sample data as the input of the model, and use the corresponding permeability in each training sample data as the expected output of the model, and obtain a permeability prediction model by using a machine learning method, including: establishing a permeability prediction model by using a support vector machine method to reflect the non-linear relationship between the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor, and shear wave attenuation factor and the permeability.

8. The permeability prediction device according to any one of claims 5-7, wherein, the acquisition module is configured to obtain the longitudinal wave velocity, shear wave velocity, density, longitudinal wave attenuation factor, and shear wave attenuation factor of the target reservoir, including: obtaining the longitudinal wave velocity, shear wave velocity, and density of the target reservoir by pre-stack three-parameter inversion; obtaining the longitudinal wave attenuation factor and shear wave attenuation factor of the target reservoir by pre-stack amplitude-versus-frequency inversion.

9. An electronic device, wherein, it includes: at least one processor and a memory; the memory stores computer execution instructions; the at least one processor executes the computer execution instructions stored in the memory, so that the at least one processor executes the permeability prediction method according to any one of claims 1-4.

10. A computer-readable storage medium, wherein, computer execution instructions are stored in the computer-readable storage medium, and when the computer execution instructions are executed by a processor, they are used to implement the permeability prediction method according to any one of claims 1-4.

Images