Random forest-based ground stress prediction method, device, electronic equipment and medium

Abstract

The present application provides a method, device, electronic device and medium for predicting geostress based on random forest, which relates to the field of geophysical exploration. The method includes: obtaining seismic data of the area to be measured; processing the seismic data to determine the Young's modulus and curvature attributes; using a trained random forest-based geostress prediction model to process the Young's modulus, curvature attributes and the product of the Young's modulus and the curvature attributes of the area to be measured to obtain the target geostress of the area to be measured. The present application constructs a random forest (RF) model with curvature attributes, Young's modulus and their product as training samples for predicting geostress. The model shows extremely high prediction accuracy, and the square correlation coefficient between the actual value and the predicted value reaches 0.95. The present application improves the accuracy of geostress prediction while ensuring the efficiency of predicting geostress.

Description

Technical Field

[0001] The present application relates to the field of geophysical exploration, and more specifically, to a method, device, electronic equipment, and medium for predicting ground stress based on random forests. Background Art

[0002] As global energy demand continues to grow, traditional oil and natural gas resources are gradually being depleted. This challenge has prompted the global energy industry to accelerate its transition to the development of unconventional oil and gas resources, with shale gas receiving particular attention due to its enormous development potential and abundant reserves. The development of unconventional oil and gas resources is not only an important part of my country's energy security strategy, but also plays a key role in the sustainable development of the economy. Traditional geostress prediction methods often have limited accuracy and struggle to capture the nonlinear relationship between complex geological factors and geostress. Current methods also have limitations in the application of machine learning to geostress prediction, such as difficulty in providing continuous stress distribution across wells. Summary of the Invention

[0003] The purpose of the embodiments of the present application is to provide a random forest-based ground stress prediction method, device, electronic device and medium to solve the above-mentioned problems existing in the prior art and improve the prediction accuracy of ground stress.

[0004] In a first aspect, a method for predicting ground stress based on random forest is provided, which may include:

[0005] Obtaining seismic data of the area to be measured;

[0006] Processing the seismic data to determine Young's modulus and curvature attributes; the Young's modulus represents the elasticity of the underground medium in the area to be measured; the curvature attribute represents the deformation and bending degree of the underground structure in the area to be measured;

[0007] The trained random forest-based geostress prediction model is used to process the Young's modulus, the curvature attribute, and the product of the Young's modulus and the curvature attribute of the area to be measured to obtain the target geostress of the area to be measured.

[0008] In one possible implementation, determining Young's modulus includes:

[0009] The seismic data are processed using a pre-stack simultaneous inversion technique to obtain Young's modulus.

[0010] In a possible implementation, the training process of the geostress prediction model includes:

[0011] Acquire historical seismic data and corresponding multiple historical well logging data in the area to be tested;

[0012] Process the historical seismic data to obtain the historical Young's modulus and historical curvature properties corresponding to each point in the gather corresponding to the historical seismic data;

[0013] For the historical Young's modulus and historical curvature attribute of any point, the historical Young's modulus, the historical curvature attribute, and the product of the historical Young's modulus and the historical curvature attribute are used as training samples;

[0014] Based on multiple sets of training samples and training labels determined by corresponding historical logging data, the configured random forest-based training model is trained to obtain a trained ground stress prediction model.

[0015] In one possible implementation, training a configured random forest-based model to be trained based on multiple sets of training samples and training labels determined from corresponding well logging data includes:

[0016] For any training sample corresponding to any historical earthquake data, the configured random forest-based training model is trained according to the training sample and the corresponding training label to obtain a decision tree;

[0017] Based on the obtained multiple decision trees, the initial geostress prediction model is determined.

[0018] In a possible implementation, after determining the initial geostress prediction model, the method further includes:

[0019] The performance of the initial geostress prediction model is tested using the test set:

[0020] If the test result of the performance test meets the preset test result, the initial geostress prediction model is determined as the geostress prediction model.

[0021] In a possible implementation, performing a performance test on the initial geostress prediction model using a test set includes:

[0022] The square correlation coefficient and / or mean absolute percentage error are used to process the predicted values ​​obtained by processing the test samples in the test set using the initial geostress prediction model and the test labels corresponding to the test samples to obtain test results.

[0023] In one possible implementation, obtaining the target geostress of the area to be measured includes:

[0024] Processing the Young's modulus, the curvature attribute, and the product of the Young's modulus and the curvature attribute of the measured area through any decision tree in the in-situ stress prediction model to obtain an initial in-situ stress;

[0025] The target ground stress is obtained by calculating the average value of the multiple initial ground stresses.

[0026] In a second aspect, a ground stress prediction device based on random forest is provided, which may include:

[0027] An acquisition unit, used for acquiring seismic data of the area to be measured;

[0028] a processing unit, configured to process the seismic data to determine Young's modulus and curvature attributes; the Young's modulus represents the elasticity of the underground medium in the area to be measured; the curvature attribute represents the deformation and bending degree of the underground structure in the area to be measured;

[0029] Furthermore, a trained random forest-based geostress prediction model is used to process the Young's modulus, curvature attribute, and the product of the Young's modulus and the curvature attribute of the area to be measured to obtain the target geostress of the area to be measured.

[0030] In a third aspect, an electronic device is provided, the electronic device including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other via the communication bus;

[0031] Memory for storing computer programs;

[0032] The processor is configured to implement any of the method steps described in the first aspect when executing a program stored in the memory.

[0033] In a fourth aspect, a computer-readable storage medium is provided, wherein a computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, any of the method steps described in the first aspect is implemented.

[0034] The present application provides a method for predicting geostress based on random forests, which includes: obtaining seismic data of the area to be measured; processing the seismic data to determine the Young's modulus and curvature attributes; using a trained random forest-based geostress prediction model to process the Young's modulus, curvature attributes and the product of the Young's modulus and the curvature attributes of the area to be measured to obtain the target geostress of the area to be measured. The present application constructs a random forest (RF) model with curvature attributes, Young's modulus and their product as training samples for predicting geostress. The model shows extremely high prediction accuracy, and the square correlation coefficient between the actual value and the predicted value reaches 0.95. The present application improves the accuracy of geostress prediction while ensuring the efficiency of predicting geostress. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without creative work.

[0036] Figure 1 A system architecture diagram for a random forest-based ground stress prediction method provided in an embodiment of the present application;

[0037] Figure 2 A schematic diagram of a flow chart of a method for predicting ground stress based on random forests provided in an embodiment of the present application;

[0038] Figure 3 A schematic diagram of a cross-sectional view provided in an embodiment of the present application;

[0039] Figure 4 A schematic diagram of the structure of a ground stress prediction device based on random forest provided in an embodiment of the present application;

[0040] Figure 5 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0041] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0042] For ease of understanding, the terms involved in the embodiments of this application are explained below:

[0043] Seismic data refers to information obtained through seismic exploration technology, which can specifically include: reflection seismic data, refraction seismic data, vertical profile seismic data, microseismic monitoring data, seismic wave data and passive source seismic data.

[0044] Young's modulus (Elastic Modulus or Young's Modulus): A fundamental parameter in material mechanics, it describes a material's ability to resist extension or compression under external forces within its elastic range. In seismic research, the velocity of seismic waves (including P- and S-waves) depends on the elastic modulus (including Young's modulus) and density of the medium. Generally, the higher the Young's modulus, the faster the seismic waves propagate through the medium. Therefore, measuring seismic wave velocity can indirectly infer the Young's modulus distribution of underground rock.

[0045] Curvature attributes: In seismic exploration, curvature analysis is a technique used to extract information about the morphology of geological interfaces, enabling a better understanding of subsurface structures. Curvature attributes refer to the spatial curvature of seismic reflection layers. They reveal variations in the geometry of strata and are useful for identifying faults, folds, and other structural features.

[0046] The random forest-based ground stress prediction method provided in the embodiment of the present application can be applied to Figure 1 In the system architecture shown in Figure 1 As shown, the system may include: a server and a terminal. The server may be a physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDNs), and basic cloud computing services such as big data and artificial intelligence platforms. The terminal may be a user equipment (UE) such as a mobile phone, smart phone, laptop, digital broadcast receiver, personal digital assistant (PDA), tablet computer (PAD), handheld device, vehicle-mounted device, wearable device, computing device or other processing device connected to a wireless modem, mobile station (MS), mobile terminal, etc. The terminal and the server may be directly or indirectly connected via wired or wireless communication, which is not limited in this application.

[0047] The terminal is used to receive the earthquake data of the area to be measured input by the user and send the earthquake data to the server.

[0048] The server is used to obtain seismic data sent by the terminal to execute a random forest-based ground stress prediction method provided in this application.

[0049] As global energy demand continues to grow, traditional oil and natural gas resources are gradually being depleted. This challenge has prompted the global energy industry to accelerate its transition to the development of unconventional oil and gas resources, with shale gas receiving particular attention due to its enormous development potential and abundant reserves. The development of unconventional oil and gas resources is not only an important component of my country's energy security strategy but also plays a key role in the sustainable development of the economy. Traditional geostress prediction methods often struggle to capture the nonlinear relationship between complex geological factors and geostress. Current machine learning applications in geostress prediction also have limitations, such as difficulty in providing continuous stress distribution across wells.

[0050] Therefore, the present application provides a ground stress prediction method based on random forest to solve the above-mentioned problems existing in the prior art and improve the prediction accuracy of ground stress.

[0051] The preferred embodiments of the present application are described below in conjunction with the drawings in the specification. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present application and are not used to limit the present application. In addition, the embodiments and features in the embodiments of the present application can be combined with each other if there is no conflict.

[0052] Figure 2 The following is a flow chart of a method for predicting ground stress based on random forests provided in an embodiment of the present application. Figure 2 As shown, the method may include:

[0053] Step S210: Process the acquired seismic data to determine Young's modulus and curvature attributes.

[0054] Specifically, a pre-stack simultaneous inversion technique is used to calculate Young's modulus from seismic waves in seismic data, and to extract curvature attributes from the seismic data.

[0055] Prestack simultaneous inversion technology uses the principle that seismic waves received by ground-based geophones, after passing through complex stratigraphic structures and rocks with varying physical properties, will cause the amplitude of the reflected waves at the same reflection point underground to vary with offset. Simply put, this characteristic of seismic reflection wave amplitude changing with offset is the foundation for determining rock formation characteristics through seismic inversion.

[0056] It should be noted that seismic data is a set of time series data, and preprocessing, such as denoising and filtering, is performed on the seismic data to improve the quality of the seismic data.

[0057] The Fatti equation is a mathematical equation that describes the propagation and reflection of seismic waves between different media inside the Earth and is used to calculate the reflection and transmission coefficients of seismic waves between two media.

[0058] When seismic waves strike an elastic interface, there's a relationship between the wave's amplitude, energy, reflection coefficient, and formation parameters. The Aki-Richards equation is a linearization of the Zoeppirtz equation, expressing the P- and S-wave velocities and densities in separate terms to facilitate inversion analysis.

[0059] The Aki-Richards equation is:

[0060] The Fatti equation is a rewrite of the Aki-Richards equation. It can control the inversion noise level and improve resolution. It also creates better correlations between variables and makes the algorithm more stable.

[0061] The Fatti equation is:

[0062] R pp (θ)=C1R p +C2R s +C3R D

[0063] Among them, R pp (θ) is the longitudinal wave reflection coefficient, which changes with the incident angle θ; C i and R j are the first influence coefficient and the second influence coefficient respectively.

[0064]

[0065]

[0066] Where γ is the ratio of the shear wave velocity to the longitudinal wave velocity, ΔVp is the change in the longitudinal wave velocity, Vp is the longitudinal wave velocity, Vs is the shear wave velocity, ΔVs is the change in the shear wave velocity, ρ is the density, and Δρ is the change in density.

[0067] First, we start from the low-frequency model and use the Fatti equation for inversion. The low-frequency model includes longitudinal wave impedance, shear wave impedance, and density model. Based on the Fatti equation, we can get R p 、R s 、R D , given the incident angle θ and the value of γ, the three coefficient terms c1, c2, and c3 can be calculated. The reflection coefficient corresponding to the angle θ can then be obtained. By convolving the wavelet and the reflection coefficient, a forward-synthesized gather can be obtained. The initial model is corrected by calculating the residuals between the synthesized gather and the actual gather. Continuous iterations are performed to obtain the final inversion result. Combining the obtained inversion results with existing geological data (such as downhole logging data), a mathematical model is established regarding the relationship between Young's modulus and other parameters (such as density and wave velocity).

[0068] The prestack simultaneous inversion technique is used to estimate the Young's modulus based on parameters such as the P-wave velocity ratio and density.

[0069] And, smoothing the seismic reflection wave data.

[0070] Calculate the gradient of the reflecting interface in the area to be measured.

[0071] The curvature is calculated using the gradient information, that is, the curvature attribute is obtained.

[0072] Step S220: Using the trained random forest-based in-situ stress prediction model, the Young's modulus, the curvature attribute, and the product of the Young's modulus and the curvature attribute of the area to be measured are processed to obtain the target in-situ stress of the area to be measured.

[0073] Before executing step S220, the training process of the geostress prediction model may include:

[0074] Obtain historical seismic data and multiple historical well logging data in the area to be tested;

[0075] The historical seismic data in the measured area is processed to obtain the historical Young's modulus and historical curvature attribute corresponding to each point in the gather corresponding to the historical seismic data; the processing process is the same as step S210. It should be noted that each point in the gather is the intersection of the horizontal and vertical axes of the coordinate system corresponding to the measured area after the axes are divided according to a certain horizontal and vertical spacing, and each intersection is called a point; therefore, the historical seismic data is a gather, which includes multiple points, and each point corresponds to a historical Young's modulus and a historical curvature attribute.

[0076] It should be noted that the area to be tested mentioned in the model training process can be at least one of an area that is exactly the same as the area to be tested in actual application, an area that partially overlaps with the area to be tested, an area adjacent to the area to be tested, and an area with the same geological characteristics as the area to be tested.

[0077] For the historical Young's modulus and historical curvature attributes of any point in the gather corresponding to the historical seismic data, the historical Young's modulus, historical curvature attribute, the product of the historical Young's modulus and the historical curvature attribute, and the historical ground stress determined by the historical logging data corresponding to the historical seismic data are combined to form a historical data set.

[0078] The product of the curvature property and Young's modulus approximately represents the formation stress.

[0079] The historical data set (multiple historical data sets) corresponding to each point is divided into a test set and a training set according to a certain ratio.

[0080] The historical Young's modulus, historical curvature attribute, and the product of historical Young's modulus and historical curvature attribute in each historical data set corresponding to the training set are used as training samples; that is, the training sample set is obtained: D b = BootstraSample(D), where b represents b subsample sets; the corresponding historical in-situ stress is used as the training label. Any subsample set can extract data from the training sample set based on the characteristics of the training label. In other words, the subsample set consists of at least one training sample.

[0081] Similarly, the historical Young's modulus, historical curvature attribute, and the product of historical Young's modulus and historical curvature attribute in each historical data set corresponding to the test set are used as test samples; the corresponding historical ground stress is used as the test label.

[0082] Afterwards, based on multiple subsample sets and corresponding training labels, the configured random forest-based training model is trained to obtain a trained initial geostress prediction model including multiple decision trees.

[0083] Specifically, for any sub-sample set, the sub-sample set and the corresponding training labels are processed to obtain a decision tree.

[0084] An initial geostress prediction model is determined based on the obtained multiple decision trees. That is, the trained initial geostress prediction model may include b decision trees.

[0085] It should be noted that if the same situation exists in multiple sub-sample sets and the corresponding sample labels are the same, only one set of the same sub-sample sets and corresponding sample labels is retained. At this time, the number of decision trees generated is less than the number of sub-sample sets; if the same situation exists in multiple sub-sample sets and the corresponding sample labels are different, at this time, the time features (or other features) of the same sub-sample sets are extracted to generate a new decision tree.

[0086] In some embodiments, the performance of the initial geostress prediction model is tested using a test set:

[0087] If the test result meets the preset test result, the initial geostress prediction model is determined as the geostress prediction model.

[0088] If the test result does not meet the preset test result, the b decision trees in the initial geostress prediction model are retrained until the test result meets the preset test result.

[0089] Furthermore, the method for performing performance testing on the initial geostress prediction model using the test set may include:

[0090] The square correlation coefficient and / or mean absolute percentage error are used to process the predicted values ​​obtained by processing the test samples in the test set using the initial geostress prediction model and the test labels corresponding to the test samples to obtain the test results.

[0091] The calculation formula of the square correlation coefficient R is as follows:

[0092]

[0093] The calculation formula for the mean absolute percentage error AAPE is as follows:

[0094]

[0095] Among them, y iactual is the test label, y ipredicted is the predicted value, N is the number of test sets involved in the test process, is the average value of the test labels, is the average of the predicted values.

[0096] Furthermore, the preset test results are generally a set standard square correlation coefficient and standard mean absolute percentage error. When using one method (square correlation coefficient or mean absolute percentage error) to test the performance of the initial geostress prediction model, if the obtained square correlation coefficient or mean absolute percentage error reaches the corresponding standard square correlation coefficient or standard mean absolute percentage error, it indicates that the test result meets the preset test result; otherwise, it indicates that the test result does not meet the preset test result.

[0097] When the performance of the initial geostress prediction model is tested using two methods (square correlation coefficient or mean absolute percentage error), when the obtained square correlation coefficient and mean absolute percentage error reach the corresponding standard square correlation coefficient and standard mean absolute percentage error, it indicates that the test results meet the preset test results; otherwise, it indicates that the test results do not meet the preset test results.

[0098] In some embodiments, combined Figure 3 As shown in the figure, we can draw a cross-plot to show the test accuracy of the initial geostress prediction model on the test set, analyze the difference between the test labels and the predicted values, and ensure the consistency of the data distribution.

[0099] Step S220 may specifically include:

[0100] Through each decision tree in the trained geostress prediction model, the Young's modulus, curvature attribute and the product of Young's modulus and curvature attribute of the measured area are processed to obtain the initial geostress;

[0101] The target ground stress is obtained by calculating the average value of the multiple initial ground stresses.

[0102] The process can be expressed as:

[0103]

[0104] Among them, T b (x) is the initial in-situ stress of the Young’s modulus, curvature attribute and the product of Young’s modulus and curvature attribute of the area to be tested by b decision trees, is the target ground stress.

[0105] In some embodiments, when multiple seismic data obtained from the area to be tested are data that change with spatial position, the seismic data at different spatial positions are labeled according to the spatial position, and the labeled seismic data are processed to obtain labeled Young's modulus and curvature attributes; the labeled Young's modulus and curvature attributes are input into a trained geostress prediction model to obtain continuous target geostress, thereby determining the distribution of the target geostress, that is, providing a continuous stress distribution between wells.

[0106] This method can use the predicted target in-situ stress to guide hydraulic fracturing development in shale reservoirs, assess areas prone to fractures, and optimize extraction strategies. Hydraulic fracturing is the primary method for increasing shale gas production. In-situ stress prediction can help determine the most suitable fracturing areas and parameters, optimize fracturing design, and enhance fracturing effectiveness, thereby increasing shale gas well productivity and, further, improving the efficiency and safety of resource extraction.

[0107] The present application provides a method for predicting geostress based on random forests, which includes: obtaining seismic data of the area to be measured; processing the seismic data to determine the Young's modulus and curvature attributes; using a trained random forest-based geostress prediction model to process the Young's modulus, curvature attributes and the product of the Young's modulus and the curvature attributes of the area to be measured to obtain the target geostress of the area to be measured. The present application constructs a random forest (RF) model with curvature attributes, Young's modulus and their product as training samples for predicting geostress. The model shows extremely high prediction accuracy, and the square correlation coefficient between the actual value and the predicted value reaches 0.95. The present application improves the accuracy of geostress prediction while ensuring the efficiency of predicting geostress.

[0108] Corresponding to the above method, the embodiment of the present application also provides a ground stress prediction device based on random forest, such as Figure 4 As shown, the device includes:

[0109] An acquisition unit 410 is used to acquire seismic data of a region to be measured;

[0110] The processing unit 420 is configured to process the seismic data to determine Young's modulus and curvature attributes; the Young's modulus represents the elasticity of the underground medium in the area to be measured; and the curvature attribute represents the deformation and curvature of the underground structure in the area to be measured;

[0111] Furthermore, a trained random forest-based geostress prediction model is used to process the Young's modulus, curvature attribute, and the product of the Young's modulus and the curvature attribute of the area to be measured to obtain the target geostress of the area to be measured.

[0112] The functions of each functional unit of the random forest-based ground stress prediction device provided in the above-mentioned embodiment of the present application can be realized through the above-mentioned method steps. Therefore, the specific working process and beneficial effects of each unit in the random forest-based ground stress prediction device provided in the embodiment of the present application will not be repeated here.

[0113] The present application also provides an electronic device, such as Figure 5 As shown, it includes a processor 510 , a communication interface 520 , a memory 530 and a communication bus 540 , wherein the processor 510 , the communication interface 520 , and the memory 530 communicate with each other via the communication bus 540 .

[0114] Memory 530, for storing computer programs;

[0115] The processor 510 is configured to execute the program stored in the memory 530 by performing the following steps:

[0116] Obtaining seismic data of the area to be measured;

[0117] Processing the seismic data to determine Young's modulus and curvature attributes; the Young's modulus represents the elasticity of the underground medium in the area to be measured; the curvature attribute represents the deformation and bending degree of the underground structure in the area to be measured;

[0118] The trained random forest-based geostress prediction model is used to process the Young's modulus, the curvature attribute, and the product of the Young's modulus and the curvature attribute of the area to be measured to obtain the target geostress of the area to be measured.

[0119] The communication bus mentioned above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. This communication bus can be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is used in the figure, but this does not mean that there is only one bus or only one type of bus.

[0120] The communication interface is used for communication between the above electronic device and other devices.

[0121] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage. Alternatively, the memory may be at least one storage device located away from the processor.

[0122] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, and discrete hardware components.

[0123] The implementation methods and beneficial effects of the various components of the electronic device in the above embodiments to solve the problems can be found in Figure 2 The various steps in the embodiment shown are implemented, therefore, the specific working process and beneficial effects of the electronic device provided by the embodiment of the present application are not repeated here.

[0124] In another embodiment provided in the present application, a computer-readable storage medium is also provided, in which instructions are stored. When the computer-readable storage medium is run on a computer, the computer executes a random forest-based ground stress prediction method described in any of the above embodiments.

[0125] In another embodiment provided by the present application, a computer program product comprising instructions is also provided, which, when executed on a computer, enables the computer to execute a random forest-based ground stress prediction method as described in any one of the above embodiments.

[0126] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the embodiments of the present application can be implemented in the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the embodiments of the present application can be implemented in the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0127] The embodiments of the present application are described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0128] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0129] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps for the function specified in one or more boxes.

[0130] Unless otherwise defined, the technical or scientific terms used in this application should have the usual meanings understood by persons of ordinary skill in the field to which the invention belongs. The words "first", "second" and similar terms used in this application do not indicate any order, quantity or importance, but are only used to distinguish different components. Words such as "include" or "comprise" mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. Words such as "connect", "couple" or "connected" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0131] Although preferred embodiments have been described in the present application, those skilled in the art may make additional changes and modifications to these embodiments once they become aware of the basic inventive concepts. Therefore, the present application is intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present application.

[0132] Obviously, those skilled in the art can make various changes and modifications to the embodiments of the present application without departing from the spirit and scope of the embodiments of the present application. Thus, if these modifications and variations of the embodiments of the present application fall within the scope of the embodiments of the present application and their equivalents, the embodiments of the present application are also intended to include these modifications and variations.

Claims

1. A ground stress prediction method based on random forest, characterized in that: The method comprises: Obtaining seismic data of the area to be measured; Processing the seismic data to determine Young's modulus and curvature attributes; the Young's modulus represents the elasticity of the underground medium in the area to be measured; the curvature attribute represents the deformation and bending degree of the underground structure in the area to be measured; The trained random forest-based geostress prediction model is used to process the Young's modulus, the curvature attribute, and the product of the Young's modulus and the curvature attribute of the area to be measured to obtain the target geostress of the area to be measured.

2. The method according to claim 1, wherein Determine Young's modulus, including: The seismic data are processed using a pre-stack simultaneous inversion technique to obtain Young's modulus.

3. The method according to claim 1, wherein The training process of the geostress prediction model includes: Acquire historical seismic data and corresponding multiple historical well logging data in the area to be tested; Process the historical seismic data to obtain the historical Young's modulus and historical curvature properties corresponding to each point in the gather corresponding to the historical seismic data; For the historical Young's modulus and historical curvature attribute of any point, the historical Young's modulus, the historical curvature attribute, and the product of the historical Young's modulus and the historical curvature attribute are used as training samples; Based on multiple sets of training samples and training labels determined by corresponding historical logging data, the configured random forest-based training model is trained to obtain a trained ground stress prediction model.

4. The method according to claim 3, wherein Based on multiple sets of training samples and training labels determined by corresponding well logging data, the configured random forest-based training model is trained, including: For any training sample corresponding to any historical earthquake data, the configured random forest-based training model is trained according to the training sample and the corresponding training label to obtain a decision tree; Based on the obtained multiple decision trees, the initial geostress prediction model is determined.

5. The method according to claim 4, wherein After determining the initial geostress prediction model, the method further includes: The performance of the initial geostress prediction model is tested using the test set: If the test result of the performance test meets the preset test result, the initial geostress prediction model is determined as the geostress prediction model.

6. The method according to claim 5, wherein The performance test of the initial geostress prediction model is performed using a test set, including: The square correlation coefficient and / or mean absolute percentage error are used to process the predicted values ​​obtained by processing the test samples in the test set using the initial geostress prediction model and the test labels corresponding to the test samples to obtain test results.

7. The method according to claim 4, wherein Obtaining the target geostress of the area to be measured includes: Processing the Young's modulus, the curvature attribute, and the product of the Young's modulus and the curvature attribute of the measured area through any decision tree in the in-situ stress prediction model to obtain an initial in-situ stress; The target ground stress is obtained by calculating the average value of the multiple initial ground stresses.

8. A ground stress prediction device based on random forest, characterized in that: The device comprises: An acquisition unit, used for acquiring seismic data of the area to be measured; a processing unit, configured to process the seismic data to determine Young's modulus and curvature attributes; the Young's modulus represents the elasticity of the underground medium in the area to be measured; the curvature attribute represents the deformation and bending degree of the underground structure in the area to be measured; Furthermore, a trained random forest-based geostress prediction model is used to process the Young's modulus, curvature attribute, and the product of the Young's modulus and the curvature attribute of the area to be measured to obtain the target geostress of the area to be measured.

9. An electronic device, characterized in that: The electronic device includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus; Memory for storing computer programs; A processor, configured to implement the method steps described in any one of claims 1 to 7 when executing a program stored in a memory.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method steps according to any one of claims 1 to 7 are implemented.

Images