Fracture numerical prediction method and system based on graph model, and terminal

Abstract

The invention discloses a fracture numerical value prediction method and system based on graph models and a terminal, and the method comprises the steps: building a graph model of each target region based on the obtained core scanning CT data and acoustic emission technology experimental data of a plurality of target regions, and building a fracture model of the corresponding target region in combination with a mechanical model; based on the fracture model of each target area, obtaining a fracture value of each target area in combination with user requirements, constructing a target training set according to the core scanning CT data, the acoustic emission technology experiment data and the fracture values of all the target areas, and training a target convolutional neural network according to the target training set to obtain a fracture prediction model; and whenever core scanning CT data and acoustic emission technology experimental data of a to-be-predicted area are obtained, processing the to-be-predicted area based on the fracture prediction model to obtain a fracture value of the to-be-predicted area. According to the invention, rapid and accurate crack value estimation can be realized.

Description

Technical Field

[0001] The present invention relates to the technical field of crack prediction, and in particular to a method, system and terminal for numerical crack prediction based on a graphical model. Background Art

[0002] During oil and gas extraction, water injection and pressurization are often used to expand cracks in order to increase the flow channels of oil and gas. The extension of cracks is likely to inadvertently activate natural faults underground, thereby damaging the original geological structure. In severe cases, it may also cause the rupture and displacement of rock bodies, and then induce other problems. Therefore, obtaining accurate values ​​of cracks can help optimize the extraction process, improve production efficiency, and reduce environmental risks, thereby providing assistance for oil and gas exploration and development, so as to achieve the effect of efficient extraction and rational use of resources.

[0003] However, before modeling oil and gas well fractures, data acquisition often requires huge costs and the actual samples obtained are relatively scarce. In addition, the mechanical simulation modeling method often still has some difficulties in simulating the interrelationships such as the intersection and bifurcation behavior of a large number of fractures. At the same time, it is more suitable for simulating large-scale fractures of hundreds of kilometers and kilometers. It is often difficult to refine at small scales of decimeters, centimeters or even millimeters, resulting in the inability to accurately and conveniently obtain the exact numerical values ​​of the fractures.

[0004] Therefore, the prior art still needs to be improved and developed. Summary of the invention

[0005] The main purpose of the present invention is to provide a method, system, terminal and computer-readable storage medium for numerical prediction of fractures based on a graphical model, aiming to solve the problems in the prior art that before modeling fractures in oil and gas wells, data acquisition often requires huge costs and the samples actually obtained are relatively scarce, and the mechanical simulation modeling method often still has some difficulties in simulating the interrelationships such as the intersection and bifurcation behavior of a large number of fractures. At the same time, it is more suitable for simulating large-scale fractures of hundreds of kilometers and kilometers, and it is often difficult to refine at small scales of decimeters, centimeters or even millimeters, resulting in the inability to accurately and conveniently obtain the accurate numerical values ​​of the fractures.

[0006] To achieve the above object, the present invention provides a numerical crack prediction method based on a graphical model, and the numerical crack prediction method based on a graphical model comprises the following steps:

[0007] Based on the core scanning CT data and acoustic emission technology experimental data of multiple target areas, a graphical model of each target area is constructed, and based on the graphical model of each target area, a fracture model of the corresponding target area is constructed in combination with a mechanical model;

[0008] Based on the fracture model of each target area, the fracture value of each target area is obtained in combination with user needs. According to the core scanning CT data, acoustic emission technology experimental data and fracture values ​​of all target areas, a target training set is constructed, and the target convolutional neural network is trained according to the target training set to obtain a fracture prediction model;

[0009] Whenever core scanning CT data and acoustic emission technology experimental data of the area to be predicted are obtained, the area to be predicted is processed based on the fracture prediction model to obtain the fracture value of the area to be predicted.

[0010] Optionally, constructing a graphical model of each target area based on the core scanning CT data and acoustic emission technology experimental data of the acquired multiple target areas specifically includes:

[0011] According to the core scanning CT data of each target area, the fracture endpoints, fracture intersections, rock formation boundary points and gaps of each target area are obtained, and based on the fracture endpoints, fracture intersections, rock formation boundary points and gaps of each target area, the nodes of each target area are generated;

[0012] According to the core scanning CT data of each target area, fractures and rock formation boundaries are obtained, and based on the fractures and rock formation boundaries of each target area, the edges of each target area are generated;

[0013] Acquire attribute information of each target area, and add the attribute information to the nodes and edges of each target area;

[0014] According to the acoustic emission technology experimental data of each target area, the type, damage and stress state of each crack are obtained, and the type, damage and stress state of each crack are added to the corresponding node;

[0015] According to the nodes and edges of each target area, a graph model of each target area is constructed.

[0016] Optionally, the acquiring the attribute information of each target area and adding the attribute information to the nodes and edges of each target area specifically includes:

[0017] Acquire attribute information of each target area, wherein the attribute information includes the locations and clustering coefficients of fracture endpoints, fracture intersections, stratum boundary points and voids of each target area, and the permeability and elastic modulus of each fracture;

[0018] The positions and clustering coefficients of the fracture endpoints, fracture intersections, stratum boundary points and voids in each target area are added to the nodes of the corresponding graph model, and the permeability and elastic modulus of each fracture are added to the edges of the corresponding graph model.

[0019] Optionally, the graphical model of each target area is combined with a mechanical model to construct a crack model of the corresponding target area, specifically including:

[0020] Define the mechanical properties of the material of the mechanical model according to the material of the target area, divide the material into grids, and construct the mechanical model;

[0021] Discontinuity is introduced into the mechanical model, crack propagation is simulated according to the objective function and the enhancement function of the crack tip, and the corresponding force coefficient estimation and crack diffusion path are obtained;

[0022] The force coefficient inference and crack diffusion path are mapped to the nodes of the corresponding graphical model to generate the corresponding crack model.

[0023] Optionally, the fracture model of each target area is based on user needs to obtain the fracture value of each target area, and a target training set is constructed according to the core scanning CT data, acoustic emission technology experimental data and fracture values ​​of all target areas, specifically including:

[0024] Based on the crack model of each target area, the crack value of each target area is obtained according to user needs;

[0025] The fracture values ​​of each target area are used as labels for the core scanning CT data and acoustic emission technology experimental data of each target area to construct training data;

[0026] Summarize all training data and construct the target training set based on all training data.

[0027] Optionally, training a target convolutional neural network according to the target training set to obtain a crack prediction model specifically includes:

[0028] According to the target training set, the target convolutional neural network is iteratively trained, and the parameters of the target convolutional neural network are correspondingly updated after each training is completed;

[0029] When the training requirements are met, the training is terminated, and the target convolutional neural network obtained from the last training is output as the crack prediction model.

[0030] Optionally, the method is based on the fracture model of each target area, and in combination with user needs, the fracture value of each target area is obtained, a target training set is constructed according to the core scanning CT data, acoustic emission technology experimental data and fracture values ​​of all target areas, and a target convolutional neural network is trained according to the target training set to obtain a fracture prediction model, and then further includes:

[0031] When the user demand changes, based on the crack model of each target area, the first crack value of each target area is obtained according to the changed user demand;

[0032] Construct the first target training set based on the core scanning CT data, acoustic emission technology experimental data and the first fracture value of all target areas;

[0033] The target convolutional neural network is trained according to the first target training set to obtain a first crack prediction model.

[0034] In addition, to achieve the above-mentioned purpose, the present invention further provides a numerical crack prediction system based on a graphical model, wherein the numerical crack prediction system based on a graphical model comprises:

[0035] A data acquisition module is used to acquire core scanning CT data and acoustic emission technology experimental data of multiple target areas, build a graphical model of each target area based on the acquired core scanning CT data and acoustic emission technology experimental data of multiple target areas, and build a fracture model of the corresponding target area based on the graphical model of each target area in combination with a mechanical model;

[0036] A model generation module is used to obtain the fracture value of each target area based on the fracture model of each target area and in combination with user needs, to construct a target training set according to the core scanning CT data, acoustic emission technology experimental data and fracture values ​​of all target areas, and to train a target convolutional neural network according to the target training set to obtain a fracture prediction model;

[0037] The result acquisition module is used to process the area to be predicted based on the crack prediction model to obtain the crack value of the area to be predicted whenever the core scanning CT data and acoustic emission technology experimental data of the area to be predicted are obtained.

[0038] In addition, to achieve the above-mentioned purpose, the present invention also provides a terminal, wherein the terminal includes: a memory, a processor, and a numerical crack prediction program based on a graphical model stored in the memory and executable on the processor, wherein the numerical crack prediction program based on the graphical model implements the steps of the numerical crack prediction method based on the graphical model as described above when executed by the processor.

[0039] In addition, to achieve the above-mentioned purpose, the present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a numerical crack prediction program based on a graphical model, and when the numerical crack prediction program based on a graphical model is executed by a processor, the steps of the numerical crack prediction method based on the graphical model as described above are implemented.

[0040] In the present invention, based on the core scanning CT data and acoustic emission technology experimental data of multiple target areas, a graphical model of each target area is constructed, and based on the graphical model of each target area, a crack model of the corresponding target area is constructed in combination with a mechanical model; based on the crack model of each target area, a crack value of each target area is obtained in combination with user needs, and a target training set is constructed according to the core scanning CT data, acoustic emission technology experimental data and crack values ​​of all target areas, and a target convolutional neural network is trained according to the target training set to obtain a crack prediction model; whenever the core scanning CT data and acoustic emission technology experimental data of the area to be predicted are obtained, the area to be predicted is processed based on the crack prediction model to obtain the crack value of the area to be predicted. The graphical model obtained by the present invention based on core scanning CT data and acoustic emission technology experimental data can be used to describe the microscopic fracture structure, while the mechanical model can be extended to a larger scale. Through this multi-scale integrated model, physical mechanisms at different levels can be covered; and the combination of the graphical model and the mechanical model realizes the global and local modeling of the entire fracture process, effectively connects the macroscopic structure with the microscopic fracture characteristics, and realizes the evaluation of the fracture behavior under multi-field coupling conditions; finally, based on the transfer learning algorithm, a fracture prediction model that meets user needs can be obtained, thereby combining the results obtained by the graphical model to achieve accurate and fast fracture numerical estimation. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a flow chart of a preferred embodiment of the numerical crack prediction method based on the graphical model of the present invention;

[0042] Figure 2 is a schematic diagram of scanning CT in the numerical crack prediction method based on the graphical model of the present invention;

[0043] Figure 3 is a schematic diagram of a graphical model in a numerical crack prediction method based on a graphical model of the present invention;

[0044] Figure 4 is a schematic diagram of a crack model in the numerical crack prediction method based on a graphical model of the present invention;

[0045] Figure 5 It is a schematic diagram of transfer learning in the numerical crack prediction method based on the graphical model of the present invention;

[0046] Figure 6 It is a schematic diagram of a technical solution in the numerical crack prediction method based on the graphical model of the present invention;

[0047] Figure 7 It is a structural diagram of a preferred embodiment of a numerical crack prediction system based on a graphical model of the present invention;

[0048] Figure 8It is a structural diagram of a preferred embodiment of the terminal of the present invention. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solution and advantages of the present invention clearer and more specific, the present invention is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0050] During oil and gas extraction, water injection and pressurization are often used to expand cracks in order to increase the flow channels of oil and gas. The extension of cracks is likely to inadvertently activate natural faults underground, thereby damaging the original geological structure. In severe cases, it may also cause the rupture and displacement of rock bodies, and then induce other problems. Therefore, obtaining accurate values ​​of cracks can help optimize the extraction process, improve production efficiency, and reduce environmental risks, thereby providing assistance for oil and gas exploration and development, so as to achieve the effect of efficient extraction and rational use of resources.

[0051] However, before modeling oil and gas well fractures, data acquisition often requires huge costs and the actual samples obtained are relatively scarce. In addition, the mechanical simulation modeling method often still has some difficulties in simulating the interrelationships such as the intersection and bifurcation behavior of a large number of fractures. At the same time, it is more suitable for simulating large-scale fractures of hundreds of kilometers and kilometers. It is often difficult to refine at small scales of decimeters, centimeters or even millimeters, resulting in the inability to accurately and conveniently obtain the exact numerical values ​​of the fractures.

[0052] In response to one or more of the above problems, the present invention constructs a graphical model of each target area based on the core scanning CT data and acoustic emission technology experimental data of multiple target areas, and constructs a crack model of the corresponding target area based on the graphical model of each target area in combination with a mechanical model; based on the crack model of each target area, the crack value of each target area is obtained in combination with user needs, and a target training set is constructed according to the core scanning CT data, acoustic emission technology experimental data and crack values ​​of all target areas, and a target convolutional neural network is trained according to the target training set to obtain a crack prediction model; whenever the core scanning CT data and acoustic emission technology experimental data of the area to be predicted are obtained, the area to be predicted is processed based on the crack prediction model to obtain the crack value of the area to be predicted.

[0053] The numerical crack prediction method based on the graphical model described in the preferred embodiment of the present invention is as follows: Figure 1 As shown, the numerical crack prediction method based on the graphical model includes the following steps:

[0054] Step S10: construct a graphical model of each target area based on the core scanning CT data and acoustic emission technology experimental data of the acquired multiple target areas, and construct a fracture model of the corresponding target area based on the graphical model of each target area in combination with the mechanical model.

[0055] Specifically, the acquisition of data before modeling fractures in a region often requires huge costs and the samples actually obtained are relatively scarce. In addition, the mechanical simulation modeling method often still has some difficulties in simulating the interrelationships such as the cross-branching behavior of a large number of fractures. At the same time, it is more suitable for simulating large-scale fractures of hundreds of kilometers and kilometers, and it is often difficult to refine at small scales of decimeters, centimeters, or even millimeters. The existence of the graph model successfully alleviates the defects of mechanical simulation modeling in the interaction, relationship and network structure between fractures and small-scale refinement.

[0056] It should be noted that core scanning CT Figure 2 As shown in the figure, the core scanning CT can obtain information such as the end points of cracks, intersection points of cracks, boundary points of rock formations, voids, cracks and rock formation boundaries. The acoustic emission technology experiment is a non-destructive testing method that evaluates the performance or structural integrity of materials by receiving and analyzing the acoustic emission signals of materials. The stress wave phenomenon caused by the rapid release of strain energy due to crack expansion, plastic deformation or phase change in the material is called acoustic emission. The corresponding data such as the type of crack, damage and stress state are obtained through experiments.

[0057] Furthermore, the method of constructing a graphical model of each target area based on the core scanning CT data and acoustic emission technology experimental data of the acquired multiple target areas specifically includes:

[0058] According to the core scanning CT data of each target area, the fracture endpoints, fracture intersections, rock formation boundary points and gaps of each target area are obtained, and based on the fracture endpoints, fracture intersections, rock formation boundary points and gaps of each target area, the nodes of each target area are generated;

[0059] According to the core scanning CT data of each target area, fractures and rock formation boundaries are obtained, and based on the fractures and rock formation boundaries of each target area, the edges of each target area are generated;

[0060] Acquire attribute information of each target area, and add the attribute information to the nodes and edges of each target area;

[0061] According to the acoustic emission technology experimental data of each target area, the type, damage and stress state of each crack are obtained, and the type, damage and stress state of each crack are added to the corresponding node;

[0062] According to the nodes and edges of each target area, a graph model of each target area is constructed.

[0063] Specifically, scanning CT data can provide three-dimensional images of the interior of rock materials to obtain the distribution and geometric characteristics of crack endpoints, crack intersections, rock boundary points, voids, cracks, rock boundary and other discontinuities. The data obtained from CT scanning can be directly used to construct the basic structure of nodes and edges in the graph model, helping to more vividly display the spatial distribution characteristics of cracks. Acoustic emission technology is mainly used to monitor the dynamic changes of materials during the generation, connection and expansion of cracks during loading, that is, data such as crack type, damage and stress state; by analyzing acoustic emission signals, it can be used to identify different types of cracks and damage, infer the stress state and potential failure mode inside the material, and help understand the mechanism of crack formation. Scanning CT provides static structural characteristics for the construction of the graph model, and acoustic emission technology promptly responds to the response of the material under dynamic changes. The combination of the two provides more complete and solid data support for the graphical model to visually display cracks.

[0064] In the present invention, when constructing a graph model, the creation of the graph model is a process of abstracting and digitizing the topological structure and geometric information of the fracture network; the present invention uses the collected geological data to construct the basic framework of the graph model, and defines the nodes and edges in the graph model, so as to simulate the initiation and expansion of some large fractures and small fractures at the centimeter and millimeter levels. Nodes are usually used to represent important spatial positions or features in the system, such as pores, endpoints or intersections of fractures, or even the boundaries of various rock formations; edges are usually used to represent the fractures themselves or to connect different nodes to determine the relationship between nodes; that is, nodes of each target area are generated according to the fracture endpoints, fracture intersections, rock formation boundary points and gaps in the target area, and edges of each target area are generated according to the fractures and rock formation boundaries in the target area, through the nodes; when the nodes and edges are obtained, the queue can form a graph model, that is, Figure 3 shown.

[0065] Then, the nodes and edges in the graph model are assigned corresponding attributes, that is, the nodes can be given location information or clustering coefficients, and the edges can be assigned information about the permeability or elastic modulus of the cracks.

[0066] After constructing the graph model, after clarifying the topological relationship between each node and edge, the maximum stress criterion or energy criterion can be used to set when and how the crack will expand. By changing the external environmental conditions, the graph model can be used to analyze the dynamic changes of the cracks, identify potential crack expansion paths, and apply them to the connectivity analysis of the cracks or the flow simulation of the crack network.

[0067] Furthermore, the acquiring of attribute information of each target area and adding the attribute information to the nodes and edges of each target area specifically includes:

[0068] Acquire attribute information of each target area, wherein the attribute information includes the locations and clustering coefficients of fracture endpoints, fracture intersections, stratum boundary points and voids of each target area, and the permeability and elastic modulus of each fracture;

[0069] The positions and clustering coefficients of the fracture endpoints, fracture intersections, stratum boundary points and voids in each target area are added to the nodes of the corresponding graph model, and the permeability and elastic modulus of each fracture are added to the edges of the corresponding graph model.

[0070] Specifically, when assigning corresponding attributes to the nodes and edges in the graph model, the acquired positions and clustering coefficients of the fracture endpoints, fracture intersections, stratum boundary points and voids are added to the corresponding nodes of the graph model, and the permeability and elastic modulus of each fracture are added to the corresponding edges of the graph model.

[0071] The creation of the graphical model not only visually displays the stress distribution within small-scale crack materials, but also obtains the relationship between different cracks, providing a certain solution to the difficult situation of mechanical simulation crack modeling when it is impossible to simply simulate small-scale cracks over a large area and visually display the intersection of a large number of cracks.

[0072] Furthermore, the graphical model of each target area is combined with the mechanical model to construct a crack model of the corresponding target area, specifically including:

[0073] Define the mechanical properties of the material of the mechanical model according to the material of the target area, divide the material into grids, and construct the mechanical model;

[0074] Discontinuity is introduced into the mechanical model, crack propagation is simulated according to the objective function and the enhancement function of the crack tip, and the corresponding force coefficient estimation and crack diffusion path are obtained;

[0075] The force coefficient inference and crack diffusion path are mapped to the nodes of the corresponding graphical model to generate the corresponding crack model.

[0076] Specifically, the mechanical model can be coupled with other physical phenomena such as fluid mechanics to simulate the expansion behavior and morphology of cracks during hydraulic fracturing, which is very important in the exploitation of unconventional areas. For example, the extended finite element method can be used through the force model to simulate the behavior and expansion process of cracks in the entire structure at the macro level.

[0077] In the present invention, first, the mechanical properties of the material need to be defined before crack modeling, and then the material is meshed: a denser mesh is arranged near the crack to capture local changes in stress and displacement, and a sparser mesh can be used in other areas; then discontinuity is introduced, and the Heaviside function and the enhancement function of the crack tip are used to simulate the expansion of the crack, wherein the Heaviside function is the objective function; after completing the above preparations, a unique parameter level set function is set to track the path of the crack; loads and boundary conditions are set, including forces or displacements applied to the structure, etc. In the process of simulating crack expansion, it is sometimes necessary to take into account the gradual increase or change of the load, so as to simulate the actual situation of the loading process in the real working environment; then the mechanical response, deformation and displacement distribution of the overall modeling can be calculated; after the simulation is completed, the corresponding force coefficients and crack diffusion paths can be extrapolated and mapped to the nodes of the graphical model to generate the following: Figure 4 The crack model shown facilitates the analysis of stress distribution of cracks in the construction area, while also updating the relationship between cracks and their impact intensity.

[0078] In the present invention, the crack model can perform a more comprehensive analysis of the system at different scales by integrating the large-scale mechanical model and the small-scale graphical model, and obtain a more comprehensive perspective. At the same time, when processing large-scale data, the message transmission mechanism of the graphical model can also further simplify the calculation process of the mechanical model and improve the model efficiency.

[0079] Step S20: Based on the fracture model of each target area, the fracture value of each target area is obtained in combination with user needs; a target training set is constructed according to the core scanning CT data, acoustic emission technology experimental data and fracture values ​​of all target areas; and a target convolutional neural network is trained according to the target training set to obtain a fracture prediction model.

[0080] Specifically, by combining the graphical model and the mechanical model, the detailed situation of the fracture extension can be observed in an all-round and multi-scale manner, such as the direction of the fracture, the distribution of stress, the stress intensity coefficient at the fracture tip, etc. These data provide a rich training data set for the subsequent transfer learning model, that is, the training of the target convolutional neural network. At the same time, it can also be combined with actual conditions, such as real-time monitoring parameters such as drilling fluid flow rate, applied pressure and formation pore pressure, to have the conditions for detailed verification and adjustment of the fracture extension simulation results, providing an accurate theoretical basis for the comprehensive performance of subsequent model training.

[0081] Furthermore, the fracture model of each target area is combined with user needs to obtain the fracture value of each target area, and a target training set is constructed according to the core scanning CT data, acoustic emission technology experimental data and fracture values ​​of all target areas, specifically including:

[0082] Based on the crack model of each target area, the crack value of each target area is obtained according to user needs;

[0083] The fracture values ​​of each target area are used as labels for the core scanning CT data and acoustic emission technology experimental data of each target area to construct training data;

[0084] Summarize all training data and construct the target training set based on all training data.

[0085] In the present invention, before training the corresponding network, a corresponding data set is constructed. During the construction, the corresponding crack values ​​are obtained from the crack model according to the corresponding user needs, so as to construct the corresponding target training set, that is, the crack values ​​are set according to the corresponding regions, and the crack values ​​are values ​​that can be obtained from the crack model.

[0086] Furthermore, the training of the target convolutional neural network according to the target training set to obtain the crack prediction model specifically includes:

[0087] According to the target training set, the target convolutional neural network is iteratively trained, and the parameters of the target convolutional neural network are correspondingly updated after each training is completed;

[0088] When the training requirements are met, the training is terminated, and the target convolutional neural network obtained from the last training is output as the crack prediction model.

[0089] like Figure 5 As shown in the figure, the transfer learning model can deeply study the complex characteristics of fracture extension from a large number of data sets generated by combining graphical models and mechanical simulations, and can automatically identify its changing trends and patterns. Through transfer learning, a good data-driven model can be established in mature areas, and then migrated to new areas with scarce data to make advance predictions of fractures. Such operations enhance the applicability of the model to various geological conditions and fracture characteristics.

[0090] In the stage of constructing the transfer learning model, the present invention uses convolutional neural networks and multiple algorithms to train the model. The training process is based on the corresponding target training set, allowing the machine learning model to conduct an in-depth analysis of the internal mechanism and evolution pattern of crack expansion. According to the specific geomechanical background, the generation, expansion and evolution of cracks during the operation of the target project are accurately grasped, involving specific parameters such as location, width, depth and directionality. Then the model trained in the source area is migrated to a new area that is different from the training area but related, and the cracks are predicted in the new area. There may be some specific features in some new areas, so some fine-tuning can be performed according to the new area. You can choose to freeze certain layers and only update the last few layers to adapt to the characteristics of the new area.

[0091] In addition, in the present invention, the training requirements are set by the user, that is, the number of training times reaches a certain requirement or the model accuracy reaches a certain requirement. Through repeated model training, the parameter settings and architecture design in the algorithm are constantly modified, and combined with cross-validation methods, the model efficiency is optimized to achieve outstanding generalization ability and accurate prediction expectations. After the model training achieves a certain effect, some comparisons and analyses can be performed using existing field crack detection data to ensure that transfer learning is effective.

[0092] Furthermore, the crack model of each target area is based on the user's needs to obtain the crack value of each target area, and a target training set is constructed according to the core scanning CT data, acoustic emission technology experimental data and crack values ​​of all target areas, and a target convolutional neural network is trained according to the target training set to obtain a crack prediction model, and then it also includes:

[0093] When the user demand changes, based on the crack model of each target area, the first crack value of each target area is obtained according to the changed user demand;

[0094] Construct the first target training set based on the core scanning CT data, acoustic emission technology experimental data and the first fracture value of all target areas;

[0095] The target convolutional neural network is trained according to the first target training set to obtain a first crack prediction model.

[0096] Specifically, since the prediction requirements for each region are different, the crack values ​​required to be obtained are different. Therefore, in the present invention, when the user requirements change, the corresponding first crack values ​​are obtained from the crack model according to the corresponding user requirements to construct the first target training set, and then the corresponding training target convolutional neural network is used to obtain the first crack prediction model. The obtained first crack prediction model can be applied to the current region to obtain the required crack values. In addition, in the present invention, after different crack prediction models are obtained, they are stored so that they can be directly applied when needed.

[0097] Step S30: Whenever the core scanning CT data and acoustic emission technology experimental data of the area to be predicted are obtained, the area to be predicted is processed based on the fracture prediction model to obtain the fracture value of the area to be predicted.

[0098] Specifically, after the corresponding fracture prediction model is obtained, whenever the core scanning CT data and acoustic emission technology experimental data of the area to be predicted are obtained, they can be input into the corresponding fracture prediction model and the fracture value required at this time can be output.

[0099] Furthermore, if Figure 6As shown, an implementation process of the present invention is performed first to collect data so as to realize the construction of a graphical model, and then the mechanical model is integrated into the graphical model to obtain a crack model, which includes a full range of crack expansion paths, expansion rates and stress field distributions. Then, corresponding data can be obtained based on the crack model to realize model migration and application, that is, a neural network or model is trained corresponding to the data to obtain a corresponding crack prediction model. The obtained crack prediction model can perform numerical crack prediction at multiple levels and angles.

[0100] The present invention constructs a graphical model of each target area based on core scanning CT data and acoustic emission technology experimental data of multiple target areas, and constructs a crack model of the corresponding target area based on the graphical model of each target area in combination with a mechanical model; based on the crack model of each target area, a crack value of each target area is obtained in combination with user needs, a target training set is constructed according to the core scanning CT data, acoustic emission technology experimental data and crack values ​​of all target areas, and a target convolutional neural network is trained according to the target training set to obtain a crack prediction model; whenever the core scanning CT data and acoustic emission technology experimental data of the area to be predicted are obtained, the area to be predicted is processed based on the crack prediction model to obtain the crack value of the area to be predicted. The graphical model obtained by the present invention based on core scanning CT data and acoustic emission technology experimental data can be used to describe the microscopic fracture structure, while the mechanical model can be extended to a larger scale. Through this multi-scale integrated model, physical mechanisms at different levels can be covered; and the combination of the graphical model and the mechanical model realizes the global and local modeling of the entire fracture process, effectively connects the macroscopic structure with the microscopic fracture characteristics, and realizes the evaluation of the fracture behavior under multi-field coupling conditions; finally, based on the transfer learning algorithm, a fracture prediction model that meets user needs can be obtained, thereby combining the results obtained by the graphical model to achieve accurate and fast fracture numerical estimation.

[0101] Furthermore, if Figure 7 As shown, based on the above-mentioned numerical crack prediction method based on the graphical model, the present invention also provides a numerical crack prediction system based on the graphical model, wherein the numerical crack prediction system based on the graphical model includes:

[0102] The data acquisition module 71 is used to construct a graphical model of each target area based on the core scanning CT data and acoustic emission technology experimental data of the acquired multiple target areas, and to construct a fracture model of the corresponding target area based on the graphical model of each target area in combination with the mechanical model;

[0103] The model generation module 72 is used to obtain the fracture value of each target area based on the fracture model of each target area and in combination with user needs, to construct a target training set according to the core scanning CT data, acoustic emission technology experimental data and fracture values ​​of all target areas, and to train the target convolutional neural network according to the target training set to obtain a fracture prediction model;

[0104] The result acquisition module 73 is used to process the area to be predicted based on the fracture prediction model to obtain the fracture value of the area to be predicted whenever the core scanning CT data and the acoustic emission technology experimental data of the area to be predicted are obtained.

[0105] Furthermore, if Figure 8 As shown, based on the above-mentioned method and system for numerical crack prediction based on the graphical model, the present invention also provides a terminal accordingly, and the terminal includes a processor 10, a memory 20 and a display 30. Figure 8 Only some components of the terminal are shown, but it should be understood that it is not required to implement all of the components shown, and more or fewer components may be implemented instead.

[0106] In some embodiments, the memory 20 may be an internal storage unit of the terminal, such as a hard disk or memory of the terminal. In other embodiments, the memory 20 may also be an external storage device of the terminal, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), etc. equipped on the terminal. Further, the memory 20 may also include both an internal storage unit of the terminal and an external storage device. The memory 20 is used to store application software and various types of data installed on the terminal, such as the program code of the installation terminal. The memory 20 may also be used to temporarily store data that has been output or is to be output. In one embodiment, a numerical crack prediction program 40 based on a graphical model is stored on the memory 20, and the numerical crack prediction program 40 based on a graphical model can be executed by the processor 10, thereby realizing the numerical crack prediction method based on a graphical model in the present invention.

[0107] In some embodiments, the processor 10 may be a central processing unit (CPU), a microprocessor or other data processing chip, used to run the program code or process data stored in the memory 20, such as executing the graphical model-based numerical crack prediction method.

[0108] In some embodiments, the display 30 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch device, etc. The display 30 is used to display information on the terminal and to display a visual user interface. The components 10-30 of the terminal communicate with each other via a system bus.

[0109] In one embodiment, when the processor 10 executes the graphical model-based crack numerical prediction program 40 in the memory 20 , the steps of the above graphical model-based crack numerical prediction method are implemented.

[0110] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a numerical crack prediction program based on a graphical model, and when the numerical crack prediction program based on a graphical model is executed by a processor, the steps of the numerical crack prediction method based on a graphical model are implemented as follows:

[0111] Based on the core scanning CT data and acoustic emission technology experimental data of multiple target areas, a graphical model of each target area is constructed, and based on the graphical model of each target area, a fracture model of the corresponding target area is constructed in combination with a mechanical model;

[0112] Based on the fracture model of each target area, the fracture value of each target area is obtained in combination with user needs. According to the core scanning CT data, acoustic emission technology experimental data and fracture values ​​of all target areas, a target training set is constructed, and the target convolutional neural network is trained according to the target training set to obtain a fracture prediction model;

[0113] Whenever core scanning CT data and acoustic emission technology experimental data of the area to be predicted are obtained, the area to be predicted is processed based on the fracture prediction model to obtain the fracture value of the area to be predicted.

[0114] The method of constructing a graphical model of each target area based on the core scanning CT data and acoustic emission technology experimental data of the acquired multiple target areas specifically includes:

[0115] According to the core scanning CT data of each target area, the fracture endpoints, fracture intersections, rock formation boundary points and gaps of each target area are obtained, and based on the fracture endpoints, fracture intersections, rock formation boundary points and gaps of each target area, the nodes of each target area are generated;

[0116] According to the core scanning CT data of each target area, fractures and rock formation boundaries are obtained, and based on the fractures and rock formation boundaries of each target area, the edges of each target area are generated;

[0117] Acquire attribute information of each target area, and add the attribute information to the nodes and edges of each target area;

[0118] According to the acoustic emission technology experimental data of each target area, the type, damage and stress state of each crack are obtained, and the type, damage and stress state of each crack are added to the corresponding node;

[0119] According to the nodes and edges of each target area, a graph model of each target area is constructed.

[0120] The step of acquiring the attribute information of each target area and adding the attribute information to the nodes and edges of each target area specifically includes:

[0121] Acquire attribute information of each target area, wherein the attribute information includes the locations and clustering coefficients of fracture endpoints, fracture intersections, stratum boundary points and voids of each target area, and the permeability and elastic modulus of each fracture;

[0122] The positions and clustering coefficients of the fracture endpoints, fracture intersections, stratum boundary points and voids in each target area are added to the nodes of the corresponding graph model, and the permeability and elastic modulus of each fracture are added to the edges of the corresponding graph model.

[0123] The method of constructing a crack model of the corresponding target area based on the graphical model of each target area in combination with the mechanical model specifically includes:

[0124] Define the mechanical properties of the material of the mechanical model according to the material of the target area, divide the material into grids, and construct the mechanical model;

[0125] Discontinuity is introduced into the mechanical model, crack propagation is simulated according to the objective function and the enhancement function of the crack tip, and the corresponding force coefficient estimation and crack diffusion path are obtained;

[0126] The force coefficient inference and crack diffusion path are mapped to the nodes of the corresponding graphical model to generate the corresponding crack model.

[0127] Among them, the fracture model of each target area is based on the user's needs to obtain the fracture value of each target area, and the target training set is constructed according to the core scanning CT data, acoustic emission technology experimental data and fracture values ​​of all target areas, which specifically includes:

[0128] Based on the crack model of each target area, the crack value of each target area is obtained according to user needs;

[0129] The fracture values ​​of each target area are used as labels for the core scanning CT data and acoustic emission technology experimental data of each target area to construct training data;

[0130] Summarize all training data and construct the target training set based on all training data.

[0131] The step of training a target convolutional neural network according to the target training set to obtain a crack prediction model specifically includes:

[0132] According to the target training set, the target convolutional neural network is iteratively trained, and the parameters of the target convolutional neural network are correspondingly updated after each training is completed;

[0133] When the training requirements are met, the training is terminated, and the target convolutional neural network obtained from the last training is output as the crack prediction model.

[0134] Among them, the fracture model based on each target area is combined with user needs to obtain the fracture value of each target area, and a target training set is constructed according to the core scanning CT data, acoustic emission technology experimental data and fracture values ​​of all target areas, and a target convolutional neural network is trained according to the target training set to obtain a fracture prediction model, and then it also includes:

[0135] When the user demand changes, based on the crack model of each target area, the first crack value of each target area is obtained according to the changed user demand;

[0136] Construct the first target training set based on the core scanning CT data, acoustic emission technology experimental data and the first fracture value of all target areas;

[0137] The target convolutional neural network is trained according to the first target training set to obtain a first crack prediction model.

[0138] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the existence of other identical elements in the process, method, article or terminal including the element.

[0139] Of course, those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing related hardware (such as a processor, a controller, etc.) through a computer program, and the program can be stored in a computer-readable storage medium that can be read by a computer, and the program can include the processes of the above-mentioned method embodiments when executed. The computer-readable storage medium can be a memory, a disk, an optical disk, etc.

[0140] It should be understood that the application of the present invention is not limited to the above examples. For ordinary technicians in this field, improvements or changes can be made based on the above description. All these improvements and changes should fall within the scope of protection of the claims attached to the present invention.

Claims

1. A numerical crack prediction method based on a graphical model, characterized in that: The numerical crack prediction method based on the graphical model includes: Based on the core scanning CT data and acoustic emission technology experimental data of multiple target areas, a graphical model of each target area is constructed, and based on the graphical model of each target area, a fracture model of the corresponding target area is constructed in combination with a mechanical model; Based on the fracture model of each target area, the fracture value of each target area is obtained in combination with user needs. According to the core scanning CT data, acoustic emission technology experimental data and fracture values ​​of all target areas, a target training set is constructed, and the target convolutional neural network is trained according to the target training set to obtain a fracture prediction model; Whenever core scanning CT data and acoustic emission technology experimental data of the area to be predicted are obtained, the area to be predicted is processed based on the fracture prediction model to obtain the fracture value of the area to be predicted.

2. The method for numerical crack prediction based on a graphical model according to claim 1, characterized in that: The method of constructing a graphical model of each target area based on the core scanning CT data and acoustic emission technology experimental data of the acquired multiple target areas specifically includes: According to the core scanning CT data of each target area, the fracture endpoints, fracture intersections, rock formation boundary points and gaps of each target area are obtained, and based on the fracture endpoints, fracture intersections, rock formation boundary points and gaps of each target area, the nodes of each target area are generated; According to the core scanning CT data of each target area, fractures and rock formation boundaries are obtained, and based on the fractures and rock formation boundaries of each target area, the edges of each target area are generated; Acquire attribute information of each target area, and add the attribute information to the nodes and edges of each target area; According to the acoustic emission technology experimental data of each target area, the type, damage and stress state of each crack are obtained, and the type, damage and stress state of each crack are added to the corresponding node; According to the nodes and edges of each target area, a graph model of each target area is constructed.

3. The method for numerical crack prediction based on a graphical model according to claim 2, characterized in that: The acquiring of attribute information of each target area and adding the attribute information to the nodes and edges of each target area specifically includes: Acquire attribute information of each target area, wherein the attribute information includes the locations and clustering coefficients of fracture endpoints, fracture intersections, stratum boundary points and voids of each target area, and the permeability and elastic modulus of each fracture; The positions and clustering coefficients of the fracture endpoints, fracture intersections, stratum boundary points and voids in each target area are added to the nodes of the corresponding graph model, and the permeability and elastic modulus of each fracture are added to the edges of the corresponding graph model.

4. The method for numerical crack prediction based on a graphical model according to claim 2, characterized in that: The method of constructing a crack model of the corresponding target area based on the graphical model of each target area in combination with the mechanical model specifically includes: Define the mechanical properties of the material of the mechanical model according to the material of the target area, divide the material into grids, and construct the mechanical model; Discontinuity is introduced into the mechanical model, crack propagation is simulated according to the objective function and the enhancement function of the crack tip, and the corresponding force coefficient estimation and crack diffusion path are obtained; The force coefficient inference and crack diffusion path are mapped to the nodes of the corresponding graphical model to generate the corresponding crack model.

5. The method for numerical crack prediction based on a graphical model according to claim 1, characterized in that: The method is based on the fracture model of each target area and combines the user's needs to obtain the fracture value of each target area, and constructs a target training set according to the core scanning CT data, acoustic emission technology experimental data and fracture values ​​of all target areas, specifically including: Based on the crack model of each target area, the crack value of each target area is obtained according to user needs; The fracture values ​​of each target area are used as labels for the core scanning CT data and acoustic emission technology experimental data of each target area to construct training data; Summarize all training data and construct the target training set based on all training data.

6. The method for numerical crack prediction based on a graphical model according to claim 1, characterized in that: The step of training a target convolutional neural network according to the target training set to obtain a crack prediction model specifically includes: According to the target training set, the target convolutional neural network is iteratively trained, and the parameters of the target convolutional neural network are correspondingly updated after each training is completed; When the training requirements are met, the training is terminated, and the target convolutional neural network obtained from the last training is output as the crack prediction model.

7. The method for numerical crack prediction based on a graphical model according to claim 1, characterized in that: The method comprises: based on the fracture model of each target area, obtaining the fracture value of each target area in combination with user needs, constructing a target training set according to the core scanning CT data, acoustic emission technology experimental data and fracture values ​​of all target areas, and training a target convolutional neural network according to the target training set to obtain a fracture prediction model, and then further comprising: When the user demand changes, based on the crack model of each target area, the first crack value of each target area is obtained according to the changed user demand; Construct the first target training set based on the core scanning CT data, acoustic emission technology experimental data and the first fracture value of all target areas; The target convolutional neural network is trained according to the first target training set to obtain a first crack prediction model.

8. A numerical crack prediction system based on a graphical model, characterized in that: The graphical model-based crack numerical prediction system includes: A data acquisition module is used to construct a graphical model of each target area based on the core scanning CT data and acoustic emission technology experimental data of multiple target areas, and to construct a fracture model of the corresponding target area based on the graphical model of each target area in combination with a mechanical model; A model generation module is used to obtain the fracture value of each target area based on the fracture model of each target area and in combination with user needs, to construct a target training set according to the core scanning CT data, acoustic emission technology experimental data and fracture values ​​of all target areas, and to train a target convolutional neural network according to the target training set to obtain a fracture prediction model; The result acquisition module is used to process the area to be predicted based on the crack prediction model to obtain the crack value of the area to be predicted whenever the core scanning CT data and acoustic emission technology experimental data of the area to be predicted are obtained.

9. A terminal, characterized in that: The terminal includes: a memory, a processor, and a numerical crack prediction program based on a graphical model stored in the memory and executable on the processor. When the numerical crack prediction program based on the graphical model is executed by the processor, the steps of the numerical crack prediction method based on the graphical model as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a numerical crack prediction program based on a graphical model, and when the numerical crack prediction program based on a graphical model is executed by a processor, the steps of the numerical crack prediction method based on a graphical model as described in any one of claims 1 to 7 are implemented.

Images