Microearthquake dry and wet event separation method and device based on Gaussian mixture model
By using Gaussian mixture models and unsupervised machine learning methods, we can quickly and accurately separate dry and wet microseismic events, solving the problem of low efficiency in removing dry and wet events in existing technologies and improving the accuracy of fracturing stimulation swept volume calculation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-24
- Publication Date
- 2026-03-24
AI Technical Summary
In existing technologies, the efficiency of removing dry and wet events from microseismic events is relatively low, which affects the accuracy of calculating the swept volume of fracturing.
An unsupervised machine learning method based on Gaussian mixture models is adopted to achieve rapid clustering and separation of dry and wet microseismic events by calculating the posterior probability of microseismic event points relative to each Gaussian component and iteratively updating the Gaussian parameters.
It enables rapid and accurate separation of dry and wet microseismic events, improving the accuracy of sweep volume calculation for fracturing.
Smart Images
Figure CN121721703A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the field of microseismic interpretation, and more particularly, to a microseismic dry and wet event separation method and device based on Gaussian mixture model. BACKGROUND
[0002] In the process of shale oil and gas hydraulic fracturing, microseismic rupture events can be divided into two categories according to different causes: one is the near-well formation rupture caused by reservoir reconstruction during hydraulic fracturing, which usually corresponds to the wave propagation area of fracturing fluid and proppant, and can be used to calculate the wave propagation volume of fracturing reconstruction, and this type of event is a wet event or an effective reconstruction event; the other is the far-well fracture activity caused by stress change and conduction during fracturing, which is not the effective reconstruction area of fracturing fluid and proppant, and affects the accuracy of wave propagation volume calculation, and needs to be removed, and this type of event is a dry event or an ineffective reconstruction event.
[0003] In the past research, the dry event is usually removed based on the time-distance relationship of microseismic events relative to the fracturing well section, but the efficiency of this removal method is low. SUMMARY
[0004] The purpose of the present application is to provide a microseismic dry and wet event separation method and device based on Gaussian mixture model, which realizes the spatial distribution of microseismic events, introduces an unsupervised machine learning method, and quickly distinguishes the effective reconstruction events (wet events) corresponding to fracturing construction and the fracture activity events (dry events) caused by stress change.
[0005] To achieve the above purpose, in the first aspect, the present application provides a microseismic dry and wet event separation method based on Gaussian mixture model, comprising:
[0006] S1: based on the microseismic detection result, a corresponding Gaussian mixture model is established, the parameters of the Gaussian mixture model are randomly initialized, and the number of mixed components is set;
[0007] S2: the posterior probability of each microseismic event point relative to each Gaussian component is calculated;
[0008] S3: based on the posterior probability of each microseismic event point relative to each Gaussian component, the Gaussian mixture distribution parameters are updated;
[0009] S4: based on the updated Gaussian mixture distribution parameters, the maximum likelihood function is calculated and whether it meets the threshold value is evaluated, if not, return to S2, if yes, execute S5;
[0010] S5: output the Gaussian mixture model after parameter update and the dry and wet event classification result, wherein the wet event is the effective reconstruction event corresponding to fracturing construction, and the dry event is the fracture activity event caused by stress change.
[0011] Optionally, in step S1, the number of mixed components is set to 2, which means that the current microseismic event distribution is divided into a group of dry events and a group of wet events.
[0012] Optionally, in step S2, the posterior probability γ of each microseismic event point relative to each Gaussian component is calculated using the following formula. nk :
[0013]
[0014] In the formula, x n π represents the spatial location of the microseismic event point. j Let be the mixing coefficient of the j-th Gaussian component. It follows a Gaussian distribution, μ j ,Σ j These are the mean and covariance matrices corresponding to the j-th Gaussian distribution, respectively, where K represents the number of mixture components, and π... k μ is the mixing coefficient of the current k-th Gaussian mixture component. k ,∑ k These are the mean and covariance matrices corresponding to the current k-th Gaussian component, respectively. K represents multiple Gaussian mixture components, and k represents one of them. γ nk Represents event point x n The posterior probability relative to the k-th Gaussian mixture component.
[0015] Optionally, in step S3, updating the Gaussian mixture distribution parameters includes:
[0016]
[0017]
[0018]
[0019] in, These are the mean, mixing coefficient, and covariance matrix of the updated Gaussian components, respectively.
[0020] Optionally, in step S4, the formula for calculating the maximum probability function is:
[0021]
[0022] In the formula, lnp(X|μ,Σ,π) is the maximum probability function value, X is all microseismic event points, and N is the number of microseismic events.
[0023] Secondly, the present invention provides an electronic device, the electronic device comprising:
[0024] At least one processor; and,
[0025] A memory communicatively connected to the at least one processor; wherein,
[0026] The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the microseismic wet and dry event separation method based on the Gaussian mixture model described in the first aspect.
[0027] Thirdly, the present invention proposes a non-transitory computer-readable storage medium storing computer instructions for causing a computer to execute the microseismic wet and dry event separation method based on the Gaussian mixture model described in the first aspect.
[0028] Fourthly, this invention proposes a microseismic wet and dry event separation device based on a Gaussian mixture model, comprising:
[0029] The data input module is used to input microseismic monitoring results;
[0030] The Gaussian mixture model building and event classification module is used to perform the following steps:
[0031] S1: Based on the microseismic detection results, establish the corresponding Gaussian mixture model, randomly initialize the parameters of the Gaussian mixture model, and set the number of mixture components;
[0032] S2: Calculate the posterior probability of each microseismic event point relative to each Gaussian component;
[0033] S3: Update the Gaussian mixture distribution parameters based on the posterior probability of each microseismic event point relative to each Gaussian component;
[0034] S4: Based on the updated Gaussian mixture distribution parameters, calculate the maximum likelihood function and evaluate whether it meets the threshold. If it does not meet the threshold, return to S2; if it does meet the threshold, execute S5.
[0035] S5: Output the updated Gaussian mixture model and the classification results of wet and dry events, where wet events are effective modification events corresponding to fracturing operations, and dry events are fracture activity events caused by changes in geostress.
[0036] Optionally, in step S1, the number of mixed components is set to 2, which means that the current microseismic event distribution is divided into a group of dry events and a group of wet events.
[0037] Optionally, in step S2, the posterior probability γ of each microseismic event point relative to each Gaussian component is calculated using the following formula. nk :
[0038]
[0039] In the formula, x n π represents the spatial location of the microseismic event point. j Let be the mixing coefficient of the j-th Gaussian component. It follows a Gaussian distribution, μ j ,Σ j These are the mean and covariance matrices corresponding to the j-th Gaussian distribution, respectively, where K represents the number of mixture components, and π... k μ is the mixing coefficient of the current k-th Gaussian mixture component. k ,Σ k These are the mean and covariance matrices corresponding to the current k-th Gaussian component, respectively. K represents multiple Gaussian mixture components, and k represents one of them. γ nk Represents event point x n The posterior probability relative to the k-th Gaussian mixture component.
[0040] The beneficial effects of this invention are as follows:
[0041] The method of this invention is based on the spatial distribution of microseismic events. It introduces an unsupervised machine learning approach and adopts a Gaussian mixture model algorithm. By evaluating the posterior probability of microseismic event points relative to each Gaussian distribution, the method uses the expectation-maximization algorithm to iteratively update the Gaussian parameters to achieve rapid clustering and separation of dry and wet microseismic events.
[0042] The system of the present invention has other features and advantages that will be apparent from or will be set forth in detail in the accompanying drawings and following detailed description, which together serve to explain the particular principles of the invention. Attached Figure Description
[0043] The above and other objects, features and advantages of the present invention will become more apparent from the accompanying drawings, in which like reference numerals generally denote like parts.
[0044] Figure 1 This is a flowchart of a microseismic wet and dry event separation method based on a Gaussian mixture model, as an example of the present invention.
[0045] Figure 2 This is a schematic diagram showing the confidence ellipse corresponding to each Gaussian component and the posterior probability of a microseismic event point relative to each Gaussian component after randomly initializing the parameters of the Gaussian mixture model in one embodiment of the present invention.
[0046] Figure 3This is a schematic diagram illustrating the changes in the position of the confidence ellipse corresponding to the model and the changes in the posterior probability of the microseismic event point relative to each Gaussian component during the process of updating the parameters of the Gaussian mixture model using the expectation-maximization algorithm in one embodiment of the present invention.
[0047] Figure 4 This is a schematic diagram of the planar distribution of the two components of the Gaussian mixture model corresponding to dry and wet events after the model parameters have converged iteratively in one embodiment of the present invention. Detailed Implementation
[0048] Gaussian mixture models are unsupervised machine learning models that assume a dataset can be linearly superimposed from multiple Gaussian distributions. They are well-suited for various data distribution types and are widely used in image segmentation, cluster analysis, and anomaly detection. Microseismic events, namely effective fracturing events (wet events) and fault activity events caused by changes in geostress (dry events), typically cluster near fracturing well sections or active faults, exhibiting significant clustering characteristics.
[0049] Therefore, this invention proposes a new method for separating dry and wet microseismic events based on the statistical laws of microseismic events. The separation of dry and wet microseismic events is regarded as an unsupervised clustering problem and is implemented based on a Gaussian mixture model.
[0050] The invention will now be described in more detail with reference to the accompanying drawings. While preferred embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that the invention will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.
[0051] Example 1
[0052] like Figure 1 As shown, this embodiment provides a method for separating wet and dry microseismic events based on a Gaussian mixture model, including:
[0053] S1: Based on the microseismic detection results, establish the corresponding Gaussian mixture model, randomly initialize the parameters of the Gaussian mixture model, and set the number of mixture components;
[0054] In this step, the number of mixed components is set to 2, which means that the current microseismic event distribution is divided into a group of dry events and a group of wet events.
[0055] S2: Calculate the posterior probability of each microseismic event point relative to each Gaussian component;
[0056] In this step, the posterior probability γ of each microseismic event point relative to each Gaussian component is calculated using the following formula. nk :
[0057]
[0058] In the formula, x n π represents the spatial location of the microseismic event point. j Let be the mixing coefficient of the j-th Gaussian component. It follows a Gaussian distribution, μ j ,Σ j These are the mean and covariance matrices corresponding to the j-th Gaussian distribution, respectively, where K represents the number of mixture components, and π... k μ is the mixing coefficient of the current k-th Gaussian mixture component. k ,∑ k These are the mean and covariance matrices corresponding to the current k-th Gaussian component, respectively. K represents multiple Gaussian mixture components, and k represents one of them. γ nk Represents event point x n The posterior probability relative to the k-th Gaussian mixture component.
[0059] S3: Update the Gaussian mixture distribution parameters based on the posterior probability of each microseismic event point relative to each Gaussian component;
[0060] In this step, updating the Gaussian mixture distribution parameters includes:
[0061]
[0062]
[0063]
[0064] in, These are the mean, mixing coefficient, and covariance matrix of the updated Gaussian components, respectively.
[0065] S4: Based on the updated Gaussian mixture distribution parameters, calculate the maximum likelihood function and evaluate whether it meets the threshold. If it does not meet the threshold, return to S2; if it does meet the threshold, execute S5.
[0066] In this step, the formula for calculating the maximum likelihood function is:
[0067]
[0068] In the formula, ln p(X|μ,∑,π) is the maximum probability function value, X is all microseismic event points, and N is the number of microseismic events.
[0069] S5: Output the updated Gaussian mixture model and the classification results of wet and dry events, where wet events are effective modification events corresponding to fracturing operations, and dry events are fracture activity events caused by changes in geostress.
[0070] This method is based on the Gaussian mixture model algorithm. By evaluating the posterior probability of microseismic event points relative to each Gaussian distribution, the Gaussian parameters are iteratively updated using the expectation-maximization algorithm, which enables rapid clustering and separation of dry and wet microseismic events.
[0071] Example 2
[0072] This embodiment provides a method for separating wet and dry microseismic events based on a Gaussian mixture model, including:
[0073] Step 1: Randomly initialize the parameters of the Gaussian mixture model and set the number of mixture components K (which represents how many groups the current microseismic event distribution is divided into; in this embodiment, it is 2 groups, i.e., K=2).
[0074] Figure 2 In this embodiment, after randomly initializing the Gaussian mixture model parameters, the confidence ellipse corresponding to each Gaussian component and the posterior probability of the microseismic event point relative to each Gaussian component are shown (yellow and purple represent two Gaussian components, and the mixed color of the two represents the posterior probability of each event point).
[0075] Step 2: For each microseismic event point, calculate its posterior probability γ relative to each Gaussian component. nk The formula is as follows:
[0076]
[0077] In the formula, x n π represents the spatial location of the microseismic event point. j The mixing coefficients for different Gaussian components are denoted as . It follows a Gaussian distribution, μ j ,Σ j Let be the mean and covariance matrix for each Gaussian distribution, respectively, and K represent the number of mixture components.
[0078] Step 3: Update the Gaussian mixture distribution parameters;
[0079]
[0080]
[0081]
[0082] Step 4: Calculate the maximum likelihood function and evaluate whether it meets the threshold, as shown in the formula below; if it does not meet the threshold, return to step 2.
[0083]
[0084] In the formula, X represents all microseismic event points, and N represents the number of microseismic events.
[0085] Figure 3 The changes in the position of the confidence ellipse corresponding to the model and the changes in the posterior probability of microseismic event points relative to each Gaussian component during the process of updating the parameters of the Gaussian mixture model using the expectation-maximization algorithm are considered. Figure 4 After the model parameters converge iteratively, the two components of the Gaussian mixture model correspond to the planar distribution of dry and wet events. It can be seen that this method can quickly distinguish between effective fracturing events (wet events) corresponding to fracturing operations and fracture activity events caused by changes in geostress (dry events).
[0086] Example 3
[0087] This embodiment provides a microseismic wet and dry event separation device based on a Gaussian mixture model, including:
[0088] The data input module is used to input microseismic monitoring results;
[0089] The Gaussian mixture model building and event classification module is used to perform the following steps:
[0090] S1: Based on the microseismic detection results, establish the corresponding Gaussian mixture model, randomly initialize the parameters of the Gaussian mixture model, and set the number of mixture components;
[0091] S2: Calculate the posterior probability of each microseismic event point relative to each Gaussian component;
[0092] S3: Update the Gaussian mixture distribution parameters based on the posterior probability of each microseismic event point relative to each Gaussian component;
[0093] S4: Based on the updated Gaussian mixture distribution parameters, calculate the maximum likelihood function and evaluate whether it meets the threshold. If it does not meet the threshold, return to S2; if it does meet the threshold, execute S5.
[0094] S5: Output the updated Gaussian mixture model and the classification results of wet and dry events, where wet events are effective modification events corresponding to fracturing operations, and dry events are fracture activity events caused by changes in geostress.
[0095] For the steps of Gaussian mixture model establishment and event classification module execution in this embodiment, please refer to Embodiment 1, which will not be repeated here.
[0096] Example 4
[0097] This embodiment provides an electronic device, the electronic device comprising:
[0098] At least one processor; and,
[0099] A memory communicatively connected to the at least one processor; wherein,
[0100] The memory stores instructions that can be executed by the at least one processor, which enables the at least one processor to perform the microseismic wet and dry event separation method based on the Gaussian mixture model described in the above embodiments.
[0101] An electronic device according to embodiments of the present disclosure includes a memory and a processor. The memory is used to store non-transitory computer-readable instructions. Specifically, the memory may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may, for example, include random access memory (RAM) and / or cache memory. The non-volatile memory may, for example, include read-only memory (ROM), hard disk, flash memory, etc.
[0102] The processor may be a central processing unit (CPU) or other form of processing unit with data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device to perform desired functions. In one embodiment of this disclosure, the processor is used to execute computer-readable instructions stored in the memory.
[0103] Those skilled in the art will understand that, in order to solve the technical problem of how to achieve a good user experience, this embodiment may also include well-known structures such as communication buses and interfaces, and these well-known structures should also be included within the protection scope of this disclosure.
[0104] For a detailed description of this embodiment, please refer to the corresponding descriptions in the foregoing embodiments, which will not be repeated here.
[0105] Example 5
[0106] This embodiment provides a non-transitory computer-readable storage medium that stores computer instructions for causing a computer to execute the microseismic wet and dry event separation method based on the Gaussian mixture model described in the above embodiment.
[0107] A computer-readable storage medium according to embodiments of the present disclosure stores non-transitory computer-readable instructions. When these non-transitory computer-readable instructions are executed by a processor, all or part of the steps of the methods described in the foregoing embodiments of the present disclosure are performed.
[0108] The aforementioned computer-readable storage media include, but are not limited to: optical storage media (e.g., CD-ROM and DVD), magneto-optical storage media (e.g., MO), magnetic storage media (e.g., magnetic tape or portable hard drive), media with built-in rewritable non-volatile memory (e.g., memory card), and media with built-in ROM (e.g., ROM cartridge).
[0109] The various embodiments of the present invention have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments.
Claims
1. A method for separating wet and dry events in microseismic events based on a Gaussian mixture model, characterized in that, include: S1: Based on the microseismic detection results, establish the corresponding Gaussian mixture model, randomly initialize the parameters of the Gaussian mixture model, and set the number of mixture components; S2: Calculate the posterior probability of each microseismic event point relative to each Gaussian component; S3: Update the Gaussian mixture distribution parameters based on the posterior probability of each microseismic event point relative to each Gaussian component; S4: Based on the updated Gaussian mixture distribution parameters, calculate the maximum likelihood function and evaluate whether it meets the threshold. If it does not meet the threshold, return to S2; if it does meet the threshold, execute S5. S5: Output the updated Gaussian mixture model and the classification results of wet and dry events, where wet events are effective modification events corresponding to fracturing operations, and dry events are fracture activity events caused by changes in geostress.
2. The method according to claim 1, characterized in that, In step S1, the number of mixed components is set to 2, which means that the current microseismic event distribution is divided into a group of dry events and a group of wet events.
3. The method according to claim 1, characterized in that, In step S2, the posterior probability γ of each microseismic event point relative to each Gaussian component is calculated using the following formula. nk : In the formula, x n π represents the spatial location of the microseismic event point. j Let be the mixing coefficient of the j-th Gaussian component. It follows a Gaussian distribution, μ j ,Σ j These are the mean and covariance matrices corresponding to the j-th Gaussian distribution, respectively, where K represents the number of mixture components, and π... k μ is the mixing coefficient of the current k-th Gaussian mixture component. k ,∑ k These are the mean and covariance matrices corresponding to the current k-th Gaussian component, respectively. K represents multiple Gaussian mixture components, and k represents one of them. γ nk Represents event point x n The posterior probability relative to the k-th Gaussian mixture component.
4. The method according to claim 3, characterized in that, In step S3, updating the Gaussian mixture distribution parameters includes: in, These are the mean, mixing coefficient, and covariance matrix of the updated Gaussian components, respectively.
5. The method according to claim 4, characterized in that, In step S4, the formula for calculating the maximum probability function is as follows: In the formula, ln p(X|μ,∑,π) is the maximum probability function value, X is all microseismic event points, and N is the number of microseismic events.
6. An electronic device, characterized in that, The electronic device includes: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the microseismic wet and dry event separation method based on the Gaussian mixture model as described in any one of claims 1-5.
7. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium stores computer instructions for causing a computer to execute the microseismic wet and dry event separation method based on the Gaussian mixture model as described in any one of claims 1-5.
8. A microseismic wet and dry event separation device based on a Gaussian mixture model, characterized in that, include: The data input module is used to input microseismic monitoring results; The Gaussian mixture model building and event classification module is used to perform the following steps: S1: Based on the microseismic detection results, establish the corresponding Gaussian mixture model, randomly initialize the parameters of the Gaussian mixture model, and set the number of mixture components; S2: Calculate the posterior probability of each microseismic event point relative to each Gaussian component; S3: Update the Gaussian mixture distribution parameters based on the posterior probability of each microseismic event point relative to each Gaussian component; S4: Based on the updated Gaussian mixture distribution parameters, calculate the maximum likelihood function and evaluate whether it meets the threshold. If it does not meet the threshold, return to S2; if it does meet the threshold, execute S5. S5: Output the updated Gaussian mixture model and the classification results of wet and dry events, where wet events are effective modification events corresponding to fracturing operations, and dry events are fracture activity events caused by changes in geostress.
9. The apparatus according to claim 8, characterized in that, In step S1, the number of mixed components is set to 2, which means that the current microseismic event distribution is divided into a group of dry events and a group of wet events.
10. The apparatus according to claim 9, characterized in that, In step S2, the posterior probability γ of each microseismic event point relative to each Gaussian component is calculated using the following formula. nk : In the formula, x n π represents the spatial location of the microseismic event point. j Let be the mixing coefficient of the j-th Gaussian component. It follows a Gaussian distribution, μ j ,∑ j These are the mean and covariance matrices corresponding to the j-th Gaussian distribution, respectively, where K represents the number of mixture components, and π... k μ is the mixing coefficient of the current k-th Gaussian mixture component. k ,Σ k These are the mean and covariance matrices corresponding to the current k-th Gaussian component, respectively. K represents multiple Gaussian mixture components, and k represents one of them. γ nk Represents event point x n The posterior probability relative to the k-th Gaussian mixture component.