Soft rock creep test system and method based on periodic dynamic load effect
Through a soft rock creep test system based on periodic dynamic load, the multi-layer attention mechanism and convolutional neural network optimization Bergus algorithm is used to solve the problem of soft rock creep test under periodic dynamic load in the existing technology. Poor versatility and low accuracy improve testing efficiency and accuracy, and enhance the safety and scientificity of oil and gas field mining.
Patent Information
- Application Number
- CN202411023957.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-29
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-07-29
AI Technical Summary
The existing soft rock creep testing methods have problems of poor generality, accuracy and low efficiency when facing periodic dynamic loads, and cannot comprehensively and accurately reflect the creep behavior of soft rocks in complex stress environments, especially in There are insufficient safety and scientificity in oil and gas field mining projects.
A soft rock creep testing system based on periodic dynamic loading was designed, including test data acquisition and normalization units, physical and mechanical properties parameter feature extraction units of underground formations, periodic dynamic loading clustering units, soft rocks Creep safety testing unit and soft rock creep parameter update unit, using multi-layer attention mechanism and convolutional neural network to optimize the Bergus algorithm, form a soft rock creep model, perform viscoelastic plasticity, damage creep and numerical simulations Test, divide influencing factors, and improve testing efficiency and accuracy.
The periodic dynamic load management of physical and mechanical properties of the underground formation parameters at different mining depths has been realized, which improves the efficiency and accuracy of soft rock creep testing, enhances the safety of underground mining of drilling platforms, and provides more scientific engineering. Design basis.
Smart Images

Figure CN118958949A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the field of testing at different mining depths, and in particular to a soft rock creep testing system and method based on periodic dynamic load action. Background Art
[0002] In the fields of oil and gas field exploitation, the creep characteristics of soft rock have an important impact on the long-term stability and safety of the project. Creep refers to the phenomenon that soft rock gradually deforms over time under continuous stress. The existing soft rock creep test methods mainly include laboratory testing and field monitoring. Although these methods reveal the creep behavior of soft rock to a certain extent, they still have many shortcomings when facing complex periodic dynamic loads.
[0003] At present, laboratory test methods such as uniaxial compression creep test and triaxial creep test usually use constant stress for loading. Although this method can simulate the creep behavior of soft rock under constant stress conditions, in actual engineering, soft rock is often subjected to periodic dynamic loads, and this dynamic stress environment cannot be accurately reflected by constant stress testing. Therefore, laboratory test methods have certain limitations in simulating creep behavior under real engineering conditions. In addition, laboratory test equipment is usually complex, cumbersome to operate, with long test cycles and high costs, which limit its wide application in actual engineering.
[0004] On the other hand, the field monitoring method records the deformation and stress changes of soft rock in real time by installing monitoring equipment in actual projects. Although this method can obtain creep data of soft rock in a real environment, field monitoring also has some disadvantages. First, field monitoring equipment is easily damaged in a complex underground environment, maintenance is difficult, and data reliability is difficult to guarantee. Secondly, the effect of periodic dynamic loads is often accompanied by environmental changes and construction interference, resulting in the monitoring data being affected by multiple factors, making it difficult to separate the creep effect of a single factor. In addition, field monitoring requires long-term continuity, a long cycle, and high costs, which is not conducive to rapid evaluation and decision-making.
[0005] In summary, the existing soft rock creep test methods have significant deficiencies when dealing with periodic dynamic loads and cannot fully and accurately reflect the creep behavior of soft rock under complex stress environments. Therefore, there is an urgent need for a soft rock creep test system and method that can simulate periodic dynamic loads and has efficient and accurate testing capabilities to meet actual engineering needs and improve the scientificity and safety of engineering design and oil and gas field exploitation. Summary of the invention
[0006] The embodiments of the present invention aim to provide a soft rock creep testing system and method based on periodic dynamic load action, and to build a soft rock creep monitoring system for underground formations of drilling platforms, so as to solve the technical problems of poor versatility, low accuracy and low efficiency of soft rock creep testing methods for underground formations of drilling platforms in the prior art.
[0007] To this end, the present invention provides a soft rock creep testing system based on periodic dynamic load, comprising:
[0008] A test data acquisition and normalization unit, used to acquire the test data at different temperatures of the underground formation of the drilling platform, and normalize the test data to form data with consistent format and integrity;
[0009] A downhole formation physical and mechanical property parameter feature extraction unit is used to complete the soft rock creep test work in the downhole formation of the drilling platform and the physical and mechanical property parameter feature extraction of the downhole formation at different mining depths by using the format consistent integrity data and the different downhole formation physical and mechanical property parameter feature extraction methods in the drilling platform control center and form abnormal features;
[0010] The periodic dynamic load clustering unit is used to analyze and cluster the soft rock creep test work of the underground formation of the drilling platform and the physical and mechanical property parameters of the underground formation at different mining depths based on the abnormal characteristics of the unit period. At the same time, it analyzes the mechanical vibration frequencies of different equipment within the periodic dynamic load at different mining depths to form analytical clustering parameters of periodic dynamic load at different mining depths;
[0011] The soft rock creep safety test unit is used to model different mining depths in the underground formation of the drilling platform to form a soft rock creep model, wherein the soft rock creep model includes a soft rock creep viscoelastic-plastic model, a soft rock damage creep model and a soft rock creep numerical simulation model; according to the soft rock creep model and the format consistent integrity data, the corresponding viscosity, elasticity, plastic deformation, damage variables, and creep behavior are tested for soft rock creep, and the factors affecting soft rock creep are divided, thereby completing the soft rock creep safety test of the underground formation of the drilling platform;
[0012] The soft rock creep parameter updating unit is used to model and identify the soft rock creep safety of the underground formation of the drilling platform using the Burgers algorithm through the soft rock creep safety test parameters of the underground formation of the drilling platform, and to update the parameters of the Burgers algorithm using the multi-layer attention mechanism and convolutional neural network optimization.
[0013] The multi-layer attention mechanism and convolutional neural network optimization are used to update the parameters of the Burgers algorithm, and the expression is:
[0014]
[0015] Among them, ε(t) represents the updated creep strain, ε0 represents the soft rock creep safety limit value, σ0 represents the stress periodic dynamic load, W1 represents the mechanical vibration frequency, η1 represents the viscosity coefficient, W2 represents the ambient temperature change variance, e represents the natural constant, Q, K, and V represent the query vector, key vector, and value vector of the attention mechanism respectively, P represents the soft rock creep influencing factor matrix, η2 represents the viscosity coefficient, t represents the unit period, f represents the parameter feature vector of the convolution input, g represents the convolution kernel, and n represents the feature dimension.
[0016] Preferably, in the above-mentioned soft rock creep test system based on periodic dynamic load, the test data acquisition and normalization unit comprises:
[0017] The test data acquisition module is used to collect the test data at different temperatures generated by different mining depths of the underground formation of the drilling platform in a unit period;
[0018] A data normalization module, used for normalizing the test data at different temperatures into corresponding physical parameter variables according to a predetermined Log transformation;
[0019] A data missing value processing module is used to process the missing values of the normalized physical parameter variables to obtain noise-free test data; the missing value processing includes but is not limited to deletion, interpolation and machine learning methods;
[0020] A periodic dynamic load source tracking module is used to track the fatigue impact degree data of different equipment from the noise-free test data, and normalize the fatigue impact degree data of different equipment according to the test depth periodic dynamic load source to obtain the equipment periodic dynamic load source information matching the test depth periodic dynamic load source;
[0021] The environment impact periodic dynamic load information analysis module is used to perform environmental impact analysis on the equipment periodic dynamic load source information to form different environmental periodic dynamic load source parameter variables as the format consistent integrity data.
[0022] Preferably, in the above-mentioned soft rock creep testing system based on periodic dynamic load, the downhole formation physical and mechanical property parameter feature extraction unit comprises:
[0023] Big data edge computing gateway adaptation module, used for big data edge computing gateway preset, adaptation mining volume and mining difficulty, and establishment of adaptation statistics table;
[0024] The drilling platform control center module is used to adjust the extraction steps and preset the extraction parameters of the physical and mechanical property parameter feature extraction method of different downhole formations in the drilling platform control center;
[0025] A module for extracting physical and mechanical property parameters of underground formations at different mining depths, which is used to extract physical and mechanical property parameters of underground formations at different mining depths periodically according to the method for extracting physical and mechanical property parameters of underground formations, the format-consistent integrity data and the adaptation statistical table, and to generate abnormal characteristics of physical and mechanical property parameters of underground formations at different mining depths;
[0026] The module for extracting the physical and mechanical property parameters characteristics of underground formations is used for the big data edge computing gateway to analyze the execution status of the soft rock creep test work, to judge and analyze the soft rock creep test work for discovering the physical and mechanical property parameters of underground formations, and to generate abnormal characteristics of the physical and mechanical property parameters of underground formations in the soft rock creep test. The abnormal characteristics of the physical and mechanical property parameters of underground formations in the soft rock creep test and the abnormal characteristics of the physical and mechanical property parameters of underground formations at different mining depths form the abnormal characteristics.
[0027] Preferably, in the above-mentioned soft rock creep testing system based on periodic dynamic load action, the periodic dynamic load action clustering unit comprises:
[0028] A downhole formation physical and mechanical property parameter analysis module is used to locate the rock formation based on the abnormal characteristics of downhole formation physical and mechanical property parameters in soft rock creep test and the abnormal characteristics of downhole formation physical and mechanical property parameters at different mining depths, including but not limited to stress analysis of downhole formation physical and mechanical property parameters, temperature analysis of downhole formation physical and mechanical property parameters, and abnormal location of downhole formation physical and mechanical property parameters;
[0029] The physical and mechanical property parameter classification and processing flow module is used to classify and form the physical and mechanical property parameter abnormal characteristics of the underground formation and the physical and mechanical property parameter information of the underground formation in the physical and mechanical property parameter abnormal characteristics of the underground formation at different mining depths, and obtain the data processing flow corresponding to the physical and mechanical property parameter information of the underground formation, and form the physical and mechanical property parameter reference benchmark of the underground formation, the coefficients to be optimized of the calculation model of the physical and mechanical property parameter of the underground formation, and the noise types;
[0030] The module of periodic dynamic load sources at different mining depths is used to detect the mechanical vibration of the entire periodic dynamic load at different mining depths, and statistically analyze the periodic dynamic load sources of equipment at different mining depths to form analytical clustering parameters of periodic dynamic load effects at different mining depths.
[0031] Preferably, in the above-mentioned soft rock creep test system based on periodic dynamic load, the soft rock creep safety test unit comprises:
[0032] Soft rock creep model management module, used to build and manage soft rock creep viscoelastic-plastic model, soft rock damage creep model and soft rock creep numerical simulation model;
[0033] The soft rock creep identification module is used to identify the soft rock creep safety of different mining depths, systems and ground stations based on the test data at different temperatures, the soft rock creep viscoelastic-plastic model, the soft rock damage creep model and the soft rock creep numerical simulation model, using the preset attention mechanism and identification algorithm to obtain the soft rock creep identification parameters;
[0034] A mining difficulty identification module is used to quantitatively evaluate the mining difficulty of underground formations at different mining depths and obtain mining difficulty identification parameters based on the soft rock creep safety identification parameters and the equipment mining difficulty periodic dynamic load source;
[0035] A soft rock creep influencing factor classification module is used to automatically complete the soft rock creep safety identification of different mining depths, systems and ground stations according to the soft rock creep identification parameters and mining difficulty identification parameters, and form a record of soft rock creep test influencing factors;
[0036] The soft rock creep process 3D visualization module is used for 3D visualization of the soft rock creep process of the system.
[0037] Preferably, in the above-mentioned soft rock creep testing system based on periodic dynamic load, the soft rock creep parameter updating unit comprises:
[0038] The record generation module is used to continuously test the working conditions of different mining depths and the periodic dynamic load sources of system equipment mining difficulty, forming records of different mining depth conditions and drilling platform mining difficulty, providing a reference for subsequent mining and soft rock creep testing;
[0039] The soft rock creep test execution module of the underground formation of the drilling platform is used to determine the soft rock creep safety at different mining depths to obtain the soft rock creep safety test parameters, evaluate the duration of different mining depths or soft rock creep levels, and evaluate the test accuracy of different mining depths for future soft rock creep tests;
[0040] The multi-layer attention mechanism and convolutional neural network optimization module are used to improve the creep safety of soft rock according to different mining depths, and the test accuracy of future soft rock creep testing work combined with different mining depths. The multi-layer attention mechanism and convolutional neural network optimization are used to update the parameters of the Burgers algorithm.
[0041] The present invention also provides a soft rock creep testing method based on periodic dynamic load, comprising the following steps:
[0042] Step S100, collecting data to be tested at different temperatures of the underground formation of the drilling platform, and normalizing the test data to form data with consistent format and integrity;
[0043] Step S200, using the format consistent integrity data and the method for extracting the physical and mechanical property parameters of different downhole formations in the drilling platform control center to complete the soft rock creep test in the downhole formation of the drilling platform and the physical and mechanical property parameter feature extraction of the downhole formations at different mining depths and form abnormal features;
[0044] Step S300: Analyze and cluster the soft rock creep test work of the underground formation of the drilling platform and the physical and mechanical property parameters of the underground formation at different mining depths according to the abnormal characteristics of the unit cycle, and analyze the mechanical vibration frequencies of different equipment under the periodic dynamic load at different mining depths to form analytical clustering parameters of the periodic dynamic load at different mining depths;
[0045] Step S400, modeling the soft rock creep test work in the underground formation of the drilling platform and the soft rock creep safety at different mining depths to form a soft rock creep model, wherein the soft rock creep model includes a soft rock creep viscoelastic-plastic model, a soft rock damage creep model and a soft rock creep numerical simulation model; performing soft rock creep tests on the corresponding viscosity, elasticity, plastic deformation, damage variables, and creep behaviors according to the soft rock creep model and the format consistent integrity data, and dividing the influencing factors of soft rock creep, thereby completing the soft rock creep safety test of the underground formation of the drilling platform;
[0046] Step S500: The soft rock creep safety of the underground formation of the drilling platform is modeled and identified using the Burgers algorithm through the soft rock creep safety test parameters of the underground formation of the drilling platform, and the parameters of the Burgers algorithm are updated using the multi-layer attention mechanism and convolutional neural network optimization.
[0047] Preferably, the above-mentioned soft rock creep test method based on periodic dynamic load is:
[0048] The steps of collecting the data to be tested at different temperatures of the underground formation of the drilling platform and normalizing the test data to form data with consistent format and integrity specifically include: collecting the data to be tested at different temperatures generated by different mining depths of the underground formation of the drilling platform in a unit period; normalizing the data to be tested at different temperatures into corresponding physical parameter variables according to a predetermined Log transformation; performing missing value processing on the normalized physical parameter variables to obtain noise-free test data; the missing value processing includes but is not limited to deletion method, interpolation method and machine learning method processing; tracking the fatigue influence degree data of different equipment from the noise-free test data, and normalizing the fatigue influence degree data of different equipment according to the periodic dynamic load source of the test depth to obtain the equipment periodic dynamic load source information matching the periodic dynamic load source of the test depth; performing environmental impact analysis on the equipment periodic dynamic load source information to form different environmental periodic dynamic load source parameter variables as the format-consistent integrity data;
[0049] The steps of using the format-consistent integrity data and the physical and mechanical property parameter feature extraction method of different downhole formations in the drilling platform control center to complete the soft rock creep test work in the downhole formation of the drilling platform and the physical and mechanical property parameter feature extraction of the downhole formation at different mining depths and form abnormal features specifically include: presetting the big data edge computing gateway, adapting the mining volume and mining difficulty, and establishing an adaptation statistical table; realizing the extraction step adjustment and extraction parameter presetting of the physical and mechanical property parameter feature extraction method of different downhole formations in the drilling platform control center; according to the physical and mechanical property parameter feature extraction method of the downhole formation, the format-consistent integrity data and the physical and mechanical property parameter feature extraction method of different downhole formations in the drilling platform control center The data and the adaptation statistical table are used to extract the characteristics of physical and mechanical property parameters of downhole formations at different mining depths periodically, and generate abnormal characteristics of physical and mechanical property parameters of downhole formations at different mining depths; the big data edge computing gateway analyzes the execution status of the soft rock creep test work, judges and analyzes the soft rock creep test work that finds the physical and mechanical property parameters of downhole formations, and generates abnormal characteristics of physical and mechanical property parameters of downhole formations in the soft rock creep test. The abnormal characteristics of physical and mechanical property parameters of downhole formations in the soft rock creep test and the abnormal characteristics of physical and mechanical property parameters of downhole formations at different mining depths form the abnormal characteristics.
[0050] Preferably, the above-mentioned soft rock creep test method based on periodic dynamic load is:
[0051] According to the abnormal characteristics of the unit cycle, the soft rock creep test work of the underground formation of the drilling platform and the physical and mechanical property parameters of the underground formation at different mining depths are analyzed and clustered, and at the same time, the mechanical vibration frequencies of different equipment under the periodic dynamic load at different mining depths are analyzed to form the analytical clustering parameters of the periodic dynamic load at different mining depths. The steps specifically include: locating the rock formation based on the abnormal characteristics of the physical and mechanical property parameters of the underground formation in the soft rock creep test and the abnormal characteristics of the physical and mechanical property parameters of the underground formation at different mining depths, including but not limited to stress analysis of the physical and mechanical property parameters of the underground formation, temperature analysis of the physical and mechanical property parameters of the underground formation, and physical and mechanical force analysis of the underground formation. The invention relates to an anomaly location of physical and mechanical property parameters; classifies and forms the anomaly characteristics of physical and mechanical property parameters of downhole formations and the physical and mechanical property parameter information of downhole formations in the anomaly characteristics of physical and mechanical property parameters of downhole formations at different mining depths, and obtains the data processing flow corresponding to the physical and mechanical property parameter information of downhole formations, forms the reference benchmark of physical and mechanical property parameters of downhole formations, the coefficients to be optimized of the calculation model of physical and mechanical property parameters of downhole formations, and the types of noise; detects the mechanical vibration of the entire periodic dynamic load at different mining depths, and statistically analyzes the sources of periodic dynamic loads of equipment at different mining depths to form analytical clustering parameters of periodic dynamic loads at different mining depths.
[0052] The soft rock creep model is formed by modeling different mining depths in the underground formation of the drilling platform, and the soft rock creep model includes a soft rock creep viscoelastic-plastic model, a soft rock damage creep model and a soft rock creep numerical simulation model; the soft rock creep test is performed on the corresponding viscosity, elasticity, plastic deformation, damage variable, and creep behavior according to the soft rock creep model and the format consistent integrity data, and the soft rock creep influencing factors are divided, and the steps of completing the soft rock creep safety test of the underground formation of the drilling platform specifically include: constructing and managing the soft rock creep viscoelastic-plastic model, the soft rock damage creep model and the soft rock creep numerical simulation model; according to the test data at different temperatures, the soft rock creep The viscoelastic-plastic model, the soft rock damage creep model and the soft rock creep numerical simulation model use a preset attention mechanism and identification algorithm to identify the soft rock creep safety of different mining depths, systems and ground stations to obtain soft rock creep identification parameters; according to the soft rock creep safety identification parameters and the periodic dynamic load source of equipment mining difficulty, the mining difficulty of underground strata at different mining depths is quantitatively evaluated and the mining difficulty identification parameters are obtained; the soft rock creep safety identification of different mining depths, systems and ground stations is automatically completed regularly or according to the soft rock creep identification parameters and mining difficulty identification parameters, and a record of influencing factors of soft rock creep test is formed; the soft rock creep process of the system is visualized in 3D.
[0053] Preferably, the above-mentioned soft rock creep test method based on periodic dynamic load is:
[0054] The steps of modeling and identifying the soft rock creep safety of the underground formation of the drilling platform using the Burgers algorithm through the soft rock creep safety test parameters of the underground formation of the drilling platform, and updating the parameters of the Burgers algorithm using the multi-layer attention mechanism and convolutional neural network optimization include:
[0055] Continuously test the working conditions of different mining depths and the periodic dynamic load sources of system equipment mining difficulty, form records of different mining depth conditions and drilling platform mining difficulty, and provide reference for subsequent mining and soft rock creep testing;
[0056] Determine the soft rock creep safety at different mining depths to obtain soft rock creep safety test parameters, evaluate the duration of different mining depths or soft rock creep levels, and evaluate the test accuracy of different mining depths for future soft rock creep tests;
[0057] According to the soft rock creep safety at different mining depths, and combined with the test accuracy of future soft rock creep testing work at different mining depths, the parameters of the Burgers algorithm are updated using a multi-layer attention mechanism and convolutional neural network optimization.
[0058] Beneficial effects:
[0059] The soft rock creep testing system and method based on periodic dynamic load action proposed in the present invention can form a universal drilling platform underground formation physical and mechanical property parameter feature extraction scheme, realize the periodic dynamic load action management of underground formation physical and mechanical property parameters at different mining depths, systematically perform physical and mechanical property parameter feature extraction of different underground formations, and test soft rock creep during formation changes, thereby ensuring high efficiency of soft rock creep testing of underground formations of drilling platforms, thereby achieving effective analysis of underground formations of drilling platforms. At the same time, the present invention optimizes the model by using a multi-layer attention mechanism and a convolutional neural network, thereby ensuring the accuracy of the Burgers algorithm for soft rock creep testing and improving the safety of underground mining on drilling platforms. BRIEF DESCRIPTION OF THE DRAWINGS
[0060] Figure 1 It is a schematic diagram of the composition of a soft rock creep testing system based on periodic dynamic load of the present invention;
[0061] Figure 2 The figure is a flow chart of a soft rock creep testing method based on periodic dynamic load of the present invention. DETAILED DESCRIPTION
[0062] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application may be combined with each other. The present application is further described in detail below in conjunction with the accompanying drawings and specific embodiments.
[0063] An embodiment of the present invention provides a soft rock creep testing system based on periodic dynamic load action, such as Figure 1 As shown, including:
[0064] The test data acquisition and normalization unit is used to integrate, normalize and process the test data at different temperatures to form data with consistent format and integrity for use by the soft rock creep test system. It is the information layer of the entire soft rock creep test system.
[0065] The downhole formation physical and mechanical property parameter feature extraction unit is used to complete soft rock creep testing and downhole formation physical and mechanical property parameter feature extraction at different mining depths based on format-consistent integrity data and different customizable downhole formation physical and mechanical property parameter data processing processes in the drilling platform control center. This unit is based on traditional downhole formation physical and mechanical property parameter feature extraction technology and realizes flexible identification of downhole formation physical and mechanical property parameter feature extraction by using format-consistent integrity data and formation change judgment algorithm.
[0066] The periodic dynamic load clustering unit is used for soft rock creep testing, physical and mechanical property parameters of underground formations at different mining depths, and analytical clustering of periodic dynamic loads at different mining depths. Through manual input and automatic integration, it completes the periodic dynamic load test at different mining depths. At the same time, it provides an analytical platform for the physical and mechanical property parameters of underground formations to complete the analytical processing of the physical and mechanical property parameters of underground formations.
[0067] The soft rock creep safety test unit is used to perform soft rock creep modeling for different mining depths and systems to form a soft rock creep model, wherein the soft rock creep model includes a soft rock creep viscoelastic-plastic model, a soft rock damage creep model and a soft rock creep numerical simulation model; based on the soft rock creep model and the format consistent integrity data, soft rock creep tests are performed on the corresponding viscosity, elasticity, plastic deformation, damage variables, and creep behaviors, and the influencing factors of soft rock creep are divided to complete the soft rock creep safety test of the system.
[0068] The soft rock creep parameter update unit is used to complete the system record generation and the soft rock creep test execution of the drilling platform underground formation through the soft rock creep safety of the system, and to update the parameters of the Burgers algorithm by using the multi-layer attention mechanism and convolutional neural network optimization for the soft rock creep problem of the system. It is the application layer of the soft rock creep information.
[0069] The multi-layer attention mechanism and convolutional neural network optimization are used to update the parameters of the Burgers algorithm, and the expression is:
[0070]
[0071] Among them, ε(t) represents the updated creep strain, ε0 represents the soft rock creep safety limit value, σ0 represents the stress periodic dynamic load, W1 represents the mechanical vibration frequency, η1 represents the viscosity coefficient, W2 represents the ambient temperature change variance, e represents the natural constant, Q, K, and V represent the query vector, key vector, and value vector of the attention mechanism respectively, P represents the soft rock creep influencing factor matrix, η2 represents the viscosity coefficient, t represents the unit period, f represents the parameter feature vector of the convolution input, g represents the convolution kernel, and n represents the feature dimension.
[0072] Among them, the physical and mechanical properties include: density, mass density of soft rock materials, usually expressed in kg / m 3 Indicates. Porosity, the ratio of the pore volume inside the soft rock to the total volume, usually expressed as a percentage. Water content, the ratio of the mass of water in the soft rock to the dry mass, usually expressed as a percentage. Compressive strength, the maximum stress when the soft rock reaches failure in a uniaxial compression test. Elastic modulus, the stress-strain relationship of the soft rock in the elastic deformation stage, indicating the stiffness of the material. Poisson's ratio, the ratio of the lateral strain to the axial strain of the soft rock under compression.
[0073] The above scheme provided in this embodiment can form a universal drilling platform downhole formation physical and mechanical property parameter feature extraction scheme, realize the periodic dynamic load management of downhole formation physical and mechanical property parameters at different mining depths, and the system's soft rock creep testing and evaluation capabilities, thereby ensuring the soft rock creep testing efficiency of downhole formations on high drilling platforms.
[0074] Preferably, the test data acquisition and normalization unit includes a test data acquisition module, a data normalization module, a data missing value processing module, a periodic dynamic load source tracking module, and an environmental impact periodic dynamic load information analysis module.
[0075] Specifically:
[0076] The test data acquisition module is used to collect the test data generated by various systems of the underground formation of the drilling platform at different temperatures in a unit period. In some schemes of the examples of the present invention, the test data acquisition module integrates different mining depth status information through wireless signals, and the status information is encapsulated using a lossless compression information structure protocol.
[0077] The data normalization module is used to format the data to be tested at different temperatures integrated into the data acquisition module to be tested according to the Log transformation, and normalize them into corresponding physical parameter variables. In the example of the present invention, the data normalization module normalizes the lossless compression information structure, and corresponds the normalized parameters to different mining depth monitoring protocols one by one according to different mining depth types to obtain their physical quantities; at the same time, the data normalization module uses the cloud to complete the normalization of the soft rock creep test work plan information, soft rock creep test work process information and other JSON formats.
[0078] The data missing value processing module is used to perform missing value processing such as deletion, interpolation, and machine learning methods on the normalized physical parameter variables to obtain standardized test data; in the example of the present invention, the deletion method includes deleting records: directly deleting records (rows) containing missing values, which is suitable for situations where the proportion of missing values is very small and randomly distributed. Deleting features: deleting features (columns) containing a large number of missing values, which is suitable for situations where the proportion of missing values of a certain feature is very high and has little impact on the analysis.
[0079] Interpolation methods include mean interpolation: replacing missing values with the mean of the feature, which is applicable to numerical data. Median interpolation: replacing missing values with the median of the feature, which is applicable to numerical data, especially when there are outliers. Mode interpolation: replacing missing values with the mode of the feature, which is applicable to categorical data. Constant interpolation: replacing missing values with specified constant values, which is applicable to special value filling in specific scenarios.
[0080] Machine learning methods include predictive model interpolation: building a predictive model (such as linear regression, decision tree, random forest, etc.) and using other features to predict missing values. Autoencoder: using a neural network model to interpolate missing values and reconstruct the data by training an autoencoder.
[0081] The periodic dynamic load source tracking module is used to select the periodic dynamic load source information of the equipment from the noise-free test data, normalize multiple types of parameters as needed, and calculate the periodic dynamic load source information of the equipment.
[0082] The environment impact periodic dynamic load information analysis module is used to analyze the environmental impact of the tracked periodic dynamic load source information to form standardized periodic dynamic load source parameters; in the example of the present invention, different state parameters at the same time are formed into different dimensional coefficients according to dimensions.
[0083] Preferably, the underground formation physical and mechanical property parameter feature extraction unit includes a big data edge computing gateway adaptation module, a drilling platform control center module, underground formation physical and mechanical property parameter feature extraction modules at different mining depths, and underground formation physical and mechanical property parameter feature extraction modules. Among them:
[0084] The big data edge computing gateway adaptation module is used for big data edge computing gateway configuration, scheduling, completing mining volume and mining difficulty supervision, and establishing adaptation statistics.
[0085] The drilling platform control center module is used to implement the characteristic extraction methods of different downhole formation physical and mechanical property parameters in the drilling platform control center, such as the preset values required for the characteristic extraction of downhole formation physical and mechanical property parameters, the extraction step adjustment and extraction parameter preset of the coefficients to be optimized and the noise types of the downhole formation physical and mechanical property parameter calculation model.
[0086] The module for extracting the characteristic parameters of underground formation physical and mechanical properties at different mining depths is used to perform periodic characteristic extraction of the physical and mechanical property parameters of underground formations at different mining depths according to preset values, coefficients to be optimized of the calculation model of the physical and mechanical property parameters of underground formations, types of noise and adaptation statistics, and generate characteristic extraction records of the physical and mechanical property parameters of underground formations at different mining depths; when the preset value range is exceeded, it is determined to be the physical and mechanical property parameters of underground formations at different mining depths; if necessary, an application for supervision of the big data edge computing gateway can also be initiated, and the characteristic extraction of the physical and mechanical property parameters of underground formations at different mining depths can be completed by further combining the adaptation statistics table.
[0087] The physical and mechanical property parameter feature extraction module of the underground formation is used for the big data edge computing gateway to analyze the execution status of the soft rock creep test work. For the soft rock creep test work in which the physical and mechanical property parameters of the underground formation are found, further judgment and analysis are performed through the big data edge computing gateway supervision, etc., and the physical and mechanical property parameter feature extraction record of the underground formation of the soft rock creep test is generated. In the example of the present invention, the physical and mechanical property parameter feature extraction of the underground formation of the soft rock creep test first analyzes and checks the execution steps of the soft rock creep test work, the rock formation positioning situation, the receiving situation, the test point alarm situation, the data record and other different inspection items in the soft rock creep test work by the noise type, and generates the receiving situation record; for the soft rock creep test work with abnormal reception, the physical and mechanical property parameter calculation model of the underground formation is optimized through the coefficient to be optimized, and the big data edge computing gateway supervision is initiated when necessary, and the physical and mechanical property parameter of the underground formation is located and analyzed through the adaptation statistical table. The physical and mechanical property parameter feature extraction record of the underground formation of the soft rock creep test is generated.
[0088] Preferably, the periodic dynamic load clustering unit includes a downhole formation physical and mechanical property parameter analysis module, a physical and mechanical property parameter classification and processing flow module, and a periodic dynamic load source module at different mining depths. Specifically:
[0089] The underground formation physical and mechanical property parameter analysis module is used for soft rock creep testing and the interaction of underground formation physical and mechanical property parameter information at different mining depths, including underground formation physical and mechanical property parameter stress analysis, underground formation physical and mechanical property parameter temperature analysis, underground formation physical and mechanical property parameter abnormal location and other information.
[0090] The physical and mechanical property parameter classification and processing flow module is used to count, analyze and refine the physical and mechanical property parameter information and data processing flow of the downhole formation, and form the reference benchmark of the physical and mechanical property parameters of the downhole formation, the coefficients to be optimized of the calculation model of the physical and mechanical property parameters of the downhole formation, and the types of noise. In the example of the present invention, the physical and mechanical property parameter classification and processing flow module statistically analyzes the downhole formation physical and mechanical property parameter disposal information recorded in the downhole formation physical and mechanical property parameter analysis module, and forms a knowledge set that is closely linked to specific business and can be used for overall analysis and judgment of the system. For typical downhole formation physical and mechanical property parameters, their knowledge set can also be summarized into the reference benchmark of downhole formation physical and mechanical property parameters and the coefficients to be optimized of the calculation model of downhole formation physical and mechanical property parameters, which are used by the downhole formation physical and mechanical property parameter feature extraction unit.
[0091] The periodic dynamic load source module of different mining depths is used to detect the mechanical vibration of the entire periodic dynamic load effect of different mining depths from commissioning, upgrading, multi-layer attention mechanism and convolutional neural network to scrapping, and at the same time, statistically analyze the periodic dynamic load sources of equipment such as the formation change state, mining difficulty, and the time node of the sudden change of the physical and mechanical property parameters of the underground formation at different mining depths. In the example of the present invention, the periodic dynamic load source module of different mining depths automatically puts different mining depths into use, upgrades, multi-layer attention mechanism and convolutional neural network, and at the same time, statistically analyzes the formation change time of different mining depths, the frequency of the physical and mechanical property parameters of the underground formation, and the reasons for the physical and mechanical property parameters of the underground formation, and regularly generates periodic dynamic load source records of different mining depths.
[0092] Preferably, the soft rock creep safety test unit includes a soft rock creep model management module, a soft rock creep identification module, a mining difficulty identification module, a soft rock creep influencing factor classification module and a soft rock creep process 3D visualization module. Specifically:
[0093] The soft rock creep model management module is used to construct and manage the soft rock creep viscoelastic plastic model, soft rock damage creep model and soft rock creep numerical simulation model. In the example of the present invention, the test points of different mining depths are classified and module to construct the model in the form of the coefficients to be optimized of the calculation model of the physical and mechanical properties parameters of the underground formation, and the relationship between each node value and the theoretical preset value is calculated to obtain the transfer node to the superior node, which is the soft rock creep model of different mining depths; the viscosity, elasticity, plastic deformation, damage variables and soft rock creep numerical simulation model are based on the soft rock creep model of different mining depths, and the system and station architecture design are integrated to obtain the coefficients to be optimized of the calculation model of the physical and mechanical properties parameters of the underground formation, which is its soft rock creep model.
[0094] The soft rock creep identification module is used to conduct a comprehensive analysis based on the test data and soft rock creep model at different temperatures, and adopt a suitable attention mechanism and identification algorithm to evaluate the soft rock creep safety of underground formations at different mining depths, systems and drilling platforms. In the example of the present invention, the soft rock creep model constructed by the soft rock creep model management module is combined with the test data at different temperatures to complete the evaluation of soft rock creep safety. The system soft rock creep safety is divided into primary creep, secondary creep, tertiary creep and quaternary creep, among which: Primary creep (level I creep) characteristics: Under low stress conditions, the creep deformation rate of soft rock is extremely slow and can basically be ignored. Applicable conditions: Applicable to those soft rock types with high strength and relatively stable structure. Mining impact: It has little impact on long-term stability and usually does not require special treatment measures.
[0095] Secondary creep (II creep) characteristics: Under moderate stress conditions, soft rock has a slow creep deformation rate, deformation mainly occurs in the initial stage, and then tends to stabilize. Applicable conditions: Applicable to those soft rock types that show some creep under moderate stress but will not quickly become unstable. Mining impact: It has a certain impact on long-term stability, and appropriate design and support measures need to be considered to control deformation.
[0096] Characteristics of level III creep: Under high stress conditions, soft rock has a fast creep deformation rate and a large deformation amount. It may stabilize within a period of time, but may continue to deform in the long term. Applicable conditions: Applicable to soft rock types that will undergo significant creep deformation under high stress. Impact of mining: Significant impact on long-term stability, requiring detailed creep analysis and effective support and reinforcement measures.
[0097] Characteristics of Grade IV creep: Under high stress conditions, soft rock has a very fast creep deformation rate and a large deformation amount, and is prone to enter the accelerated creep stage, which may eventually lead to rock mass destruction. Applicable conditions: Applicable to soft rock types that will quickly become unstable under high stress. Mining impact: It has a great impact on long-term stability, and strong support and reinforcement measures must be taken, and continuous monitoring and maintenance must be carried out.
[0098] The mining difficulty identification module is used to quantitatively evaluate the mining difficulty of drilling platforms at different mining depths, systems and drilling platform downhole formations based on the periodic dynamic load sources of the mining difficulty of the system equipment.
[0099] The soft rock creep influencing factor classification module is used to automatically organize and complete system supervision according to the soft rock creep assessment and mining difficulty identification parameters, and form a soft rock creep test influencing factor record. In the example of the present invention, when the soft rock creep influencing factor classification module detects changes in the system soft rock creep safety or the drilling platform mining difficulty, it will automatically initiate corresponding supervision and form a soft rock creep test record in combination with the soft rock creep model.
[0100] The soft rock creep process 3D visualization module is used to visualize the soft rock creep process of the system in 3D in the form of graphics and tables.
[0101] Preferably, the soft rock creep parameter updating unit includes a record generation module, a drilling platform downhole formation soft rock creep test execution module, and a multi-layer attention mechanism and convolutional neural network optimization module. More specifically:
[0102] The record generation module is used to continuously test the working status of different mining depths and the periodic dynamic load source of the system's equipment mining difficulty, forming a longitudinal record generation of different mining depth status and drilling platform mining difficulty.
[0103] The drilling platform underground formation soft rock creep test execution module is used to determine the soft rock creep safety at different mining depths, evaluate the duration of different mining depths or soft rock creep levels, and evaluate the test accuracy of future soft rock creep tests. In the present invention, an information-driven drilling platform underground formation soft rock creep test execution technology is adopted, which does not require accurate physical models and prior knowledge of different mining depths and systems. It is based on integrated historical information and uses information analysis and processing methods such as machine learning and neural networks to mine implicit information for evaluation.
[0104] The multi-layer attention mechanism and convolutional neural network optimization module are used to identify parameters according to the system soft rock creep and mining difficulty, and at the same time, combined with the execution of the soft rock creep test of the underground formation of the drilling platform, a comprehensive judgment is given to give the multi-layer attention mechanism and convolutional neural network optimization. In the example of the present invention, the multi-layer attention mechanism and convolutional neural network optimization module collects the system soft rock creep information generated by the soft rock creep safety test unit and the execution parameters of the soft rock creep test of the underground formation of the drilling platform per unit period, and combines the preset noise types to give the multi-layer attention mechanism and convolutional neural network decision suggestions. At the same time, according to the record generation and the parameters of the execution of the soft rock creep test of the underground formation of the drilling platform, the physical and mechanical properties of the underground formation that may appear are warned in advance, and maintenance suggestions are provided.
[0105] In some embodiments of the present invention, a soft rock creep test method based on periodic dynamic load is also provided. Figure 2 As shown, the following steps may be included:
[0106] Step S100, collecting data to be tested at different temperatures of the underground formation of the drilling platform, and normalizing the test data to form data with consistent format and integrity;
[0107] The resulting consistent format integrity data is used by the soft rock creep testing system.
[0108] Step S200, using the format consistent integrity data and the method for extracting the physical and mechanical property parameters of different downhole formations in the drilling platform control center to complete the soft rock creep test in the downhole formation of the drilling platform and the physical and mechanical property parameter feature extraction of the downhole formations at different mining depths and form abnormal features;
[0109] Based on the traditional extraction technology of physical and mechanical property parameters of underground formations, flexible identification of physical and mechanical property parameters of underground formations is achieved by using format-consistent integrity data and formation change judgment algorithm.
[0110] Step S300: Analyze and cluster the soft rock creep test work of the underground formation of the drilling platform and the physical and mechanical property parameters of the underground formation at different mining depths according to the abnormal characteristics of the unit cycle, and analyze the mechanical vibration frequencies of different equipment under the periodic dynamic load at different mining depths to form analytical clustering parameters of the periodic dynamic load at different mining depths;
[0111] Through manual input and automatic integration, the periodic dynamic load test at different mining depths is completed. At the same time, a platform for analyzing the physical and mechanical properties of underground formations is provided to complete the analysis and processing of the physical and mechanical properties of underground formations.
[0112] Step S400, modeling the soft rock creep test work in the underground formation of the drilling platform and the soft rock creep safety at different mining depths to form a soft rock creep model, wherein the soft rock creep model includes a soft rock creep viscoelastic plastic model, a soft rock damage creep model and a soft rock creep numerical simulation model;
[0113] According to the soft rock creep model and the format consistent integrity data, the corresponding viscosity, elasticity, plastic deformation, damage variables, and creep behavior are tested for soft rock creep, and the influencing factors of soft rock creep are divided, thereby completing the soft rock creep safety test of the underground formation of the drilling platform.
[0114] Step S500: The soft rock creep safety of the underground formation of the drilling platform is modeled and identified using the Burgers algorithm through the soft rock creep safety test parameters of the underground formation of the drilling platform, and the parameters of the Burgers algorithm are updated using the multi-layer attention mechanism and convolutional neural network optimization.
[0115] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical periodic dynamic load sources therein may be replaced by equivalent ones. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A soft rock creep test system based on periodic dynamic load, characterized in that: include: A test data acquisition and normalization unit, used to acquire the test data at different temperatures of the underground formation of the drilling platform, and normalize the test data to form data with consistent format and integrity; A downhole formation physical and mechanical property parameter feature extraction unit is used to complete the soft rock creep test work in the downhole formation of the drilling platform and the physical and mechanical property parameter feature extraction of the downhole formation at different mining depths by using the format consistent integrity data and the physical and mechanical property parameter feature extraction method of different downhole formations in the drilling platform control center and form abnormal features, wherein the physical and mechanical property parameters include: density, porosity, water content, compressive strength, elastic modulus, Poisson's ratio; The periodic dynamic load clustering unit is used to analyze and cluster the soft rock creep test work of the underground formation of the drilling platform and the physical and mechanical property parameters of the underground formation at different mining depths based on the abnormal characteristics of the unit period. At the same time, it analyzes the mechanical vibration frequencies of different equipment within the periodic dynamic load at different mining depths to form analytical clustering parameters of periodic dynamic load at different mining depths; The soft rock creep safety test unit is used to model different mining depths in the underground formation of the drilling platform to form a soft rock creep model, wherein the soft rock creep model includes a soft rock creep viscoelastic-plastic model, a soft rock damage creep model and a soft rock creep numerical simulation model; according to the soft rock creep model and the format consistent integrity data, the corresponding viscosity, elasticity, plastic deformation, damage variables, and creep behavior are tested for soft rock creep, and the factors affecting soft rock creep are divided, thereby completing the soft rock creep safety test of the underground formation of the drilling platform; The soft rock creep parameter updating unit is used to model the soft rock creep safety of the underground formation of the drilling platform using the Burgers algorithm and identify the soft rock creep safety through the soft rock creep safety test parameters of the underground formation of the drilling platform, and to update the parameters of the Burgers algorithm using the multi-layer attention mechanism and convolutional neural network optimization; The multi-layer attention mechanism and convolutional neural network optimization are used to update the parameters of the Burgers algorithm, and the expression is: Among them, ε(t) represents the updated creep strain, ε0 represents the soft rock creep safety limit value, σ0 represents the stress periodic dynamic load, W1 represents the mechanical vibration frequency, η1 represents the viscosity coefficient, W2 represents the ambient temperature change variance, e represents the natural constant, Q, K, and V represent the query vector, key vector, and value vector of the attention mechanism respectively, P represents the soft rock creep influencing factor matrix, η2 represents the viscosity coefficient, t represents the unit period, f represents the parameter feature vector of the convolution input, g represents the convolution kernel, and n represents the feature dimension.
2. A soft rock creep testing system based on periodic dynamic load according to claim 1, characterized in that: The test data acquisition and normalization unit comprises: The test data acquisition module is used to collect the test data at different temperatures generated by different mining depths of the underground formation of the drilling platform in a unit period; A data normalization module, used for normalizing the test data at different temperatures into corresponding physical parameter variables according to a predetermined Log transformation; Data missing value processing module, used to process the missing values of normalized physical parameters to obtain noise-free test data; missing value processing includes deletion method, interpolation method and machine learning method processing; A periodic dynamic load source tracking module is used to track the fatigue impact degree data of different equipment from the noise-free test data, and normalize the fatigue impact degree data of different equipment according to the test depth periodic dynamic load source to obtain the equipment periodic dynamic load source information matching the test depth periodic dynamic load source; The environment impact periodic dynamic load information analysis module is used to perform environmental impact analysis on the equipment periodic dynamic load source information to form different environmental periodic dynamic load source parameter variables as the format consistent integrity data.
3. The soft rock creep testing system based on periodic dynamic load according to claim 1 is characterized in that: The underground formation physical and mechanical property parameter feature extraction unit includes: a big data edge computing gateway adaptation module, which is used for the big data edge computing gateway to preset and adapt the mining volume and mining difficulty, and establish an adaptation statistical table; The drilling platform control center module is used to adjust the extraction steps and preset the extraction parameters of the physical and mechanical property parameter feature extraction method of different downhole formations in the drilling platform control center; A module for extracting physical and mechanical property parameters of underground formations at different mining depths, which is used to extract physical and mechanical property parameters of underground formations at different mining depths periodically according to the method for extracting physical and mechanical property parameters of underground formations, the format-consistent integrity data and the adaptation statistical table, and to generate abnormal characteristics of physical and mechanical property parameters of underground formations at different mining depths; The module for extracting the physical and mechanical property parameters characteristics of underground formations is used for the big data edge computing gateway to analyze the execution status of the soft rock creep test work, to judge and analyze the soft rock creep test work for discovering the physical and mechanical property parameters of underground formations, and to generate abnormal characteristics of the physical and mechanical property parameters of underground formations in the soft rock creep test. The abnormal characteristics of the physical and mechanical property parameters of underground formations in the soft rock creep test and the abnormal characteristics of the physical and mechanical property parameters of underground formations at different mining depths form the abnormal characteristics.
4. The soft rock creep testing system based on periodic dynamic load according to claim 1 is characterized in that: The periodic dynamic load action clustering unit comprises: The downhole formation physical and mechanical property parameter analysis module is used to locate the rock formation based on the abnormal characteristics of the downhole formation physical and mechanical property parameters in the soft rock creep test and the abnormal characteristics of the downhole formation physical and mechanical property parameters at different mining depths, including stress analysis of downhole formation physical and mechanical property parameters, temperature analysis of downhole formation physical and mechanical property parameters, and abnormal location of downhole formation physical and mechanical property parameters; The physical and mechanical property parameter classification and processing flow module is used to classify and form the physical and mechanical property parameter abnormal characteristics of the underground formation and the physical and mechanical property parameter information of the underground formation in the physical and mechanical property parameter abnormal characteristics of the underground formation at different mining depths, and obtain the data processing flow corresponding to the physical and mechanical property parameter information of the underground formation, and form the physical and mechanical property parameter reference benchmark of the underground formation, the coefficients to be optimized of the calculation model of the physical and mechanical property parameter of the underground formation, and the noise types; The module of periodic dynamic load sources at different mining depths is used to detect the mechanical vibration of the entire periodic dynamic load at different mining depths, and statistically analyze the periodic dynamic load sources of equipment at different mining depths to form analytical clustering parameters of periodic dynamic load effects at different mining depths.
5. The soft rock creep testing system based on periodic dynamic load according to claim 1 is characterized in that: The soft rock creep safety test unit includes: a soft rock creep model management module, which is used to construct and manage soft rock creep viscoelastic-plastic models, soft rock damage creep models and soft rock creep numerical simulation models; The soft rock creep identification module is used to identify the soft rock creep safety of different mining depths, systems and ground stations based on the test data at different temperatures, the soft rock creep viscoelastic-plastic model, the soft rock damage creep model and the soft rock creep numerical simulation model, using the preset attention mechanism and identification algorithm to obtain the soft rock creep identification parameters; A mining difficulty identification module is used to quantitatively evaluate the mining difficulty of underground formations at different mining depths and obtain mining difficulty identification parameters based on the soft rock creep safety identification parameters and the equipment mining difficulty periodic dynamic load source; The module for classifying influencing factors of soft rock creep is used to automatically complete the soft rock creep safety identification of different mining depths, systems and ground stations according to the soft rock creep identification parameters and mining difficulty identification parameters, and form a record of influencing factors of soft rock creep test; the module for 3D visualization of soft rock creep process is used to visualize the soft rock creep process of the system in 3D.
6. The soft rock creep testing system based on periodic dynamic load according to claim 1 is characterized in that: The soft rock creep parameter updating unit includes: a record generation module, which is used to continuously test the working status of different mining depths and the periodic dynamic load source of the mining difficulty of the system equipment, and form records of different mining depths and the mining difficulty of the drilling platform, so as to provide a reference basis for subsequent mining and soft rock creep testing; The soft rock creep test execution module of the underground formation of the drilling platform is used to determine the soft rock creep safety at different mining depths to obtain the soft rock creep safety test parameters, evaluate the duration of different mining depths or soft rock creep levels, and evaluate the test accuracy of different mining depths for future soft rock creep tests; The multi-layer attention mechanism and convolutional neural network optimization module are used to improve the creep safety of soft rock according to different mining depths, and the test accuracy of future soft rock creep testing work combined with different mining depths. The multi-layer attention mechanism and convolutional neural network optimization are used to update the parameters of the Burgers algorithm.
7. A soft rock creep testing system based on periodic dynamic load according to any one of claims 1 to 6, characterized in that: The operation process of the system is realized through a soft rock creep test method based on periodic dynamic load, which includes the following steps: Step S100, collecting data to be tested at different temperatures of the underground formation of the drilling platform, and normalizing the test data to form data with consistent format and integrity; Step S200, using the format consistent integrity data and the method for extracting the physical and mechanical property parameters of different downhole formations in the drilling platform control center to complete the soft rock creep test in the downhole formation of the drilling platform and the physical and mechanical property parameter feature extraction of the downhole formations at different mining depths and form abnormal features; Step S300: Analyze and cluster the soft rock creep test work of the underground formation of the drilling platform and the physical and mechanical property parameters of the underground formation at different mining depths according to the abnormal characteristics of the unit cycle, and analyze the mechanical vibration frequencies of different equipment under the periodic dynamic load at different mining depths to form analytical clustering parameters of the periodic dynamic load at different mining depths; Step S400, modeling the soft rock creep test work in the underground formation of the drilling platform and the soft rock creep safety at different mining depths to form a soft rock creep model, wherein the soft rock creep model includes a soft rock creep viscoelastic plastic model, a soft rock damage creep model and a soft rock creep numerical simulation model; Step S500: Perform soft rock creep test on corresponding viscosity, elasticity, plastic deformation, damage variable and creep behavior according to the soft rock creep model and the format consistent integrity data, and divide the influencing factors of soft rock creep, so as to complete the soft rock creep safety test of the downhole formation of the drilling platform; use the Burgers algorithm to model the soft rock creep safety of the downhole formation of the drilling platform and identify the soft rock creep safety through the soft rock creep safety test parameters of the downhole formation of the drilling platform, and use the multi-layer attention mechanism and convolutional neural network optimization to update the parameters of the Burgers algorithm.
8. The soft rock creep testing method based on periodic dynamic load according to claim 7 is characterized in that: The steps of collecting the test data at different temperatures of the underground formation of the drilling platform and normalizing the test data to form data with consistent format and integrity specifically include: collecting the test data at different temperatures generated by different mining depths of the underground formation of the drilling platform in a unit period; Normalizing the test data at different temperatures into corresponding physical parameter variables according to a predetermined Log transformation; performing missing value processing on the normalized physical parameter variables to obtain noise-free test data; Missing value processing includes deletion, interpolation and machine learning methods; tracking the fatigue influence data of different equipment from the noise-free test data, normalizing the fatigue influence data of different equipment according to the test depth periodic dynamic load source to obtain the equipment periodic dynamic load source information matching the test depth periodic dynamic load source; Performing environmental impact analysis on the periodic dynamic load source information of the equipment to form periodic dynamic load source parameter variables in different environments as the format-consistent integrity data; The steps of using the format consistent integrity data and the physical and mechanical property parameter feature extraction method of different downhole formations in the drilling platform control center to complete the soft rock creep test work in the downhole formation of the drilling platform and the physical and mechanical property parameter feature extraction of the downhole formations at different mining depths and form abnormal features specifically include: presetting the big data edge computing gateway, adapting the mining volume and mining difficulty, and establishing an adaptation statistical table; Realize the adjustment of extraction steps and preset extraction parameters of the feature extraction method of physical and mechanical property parameters of different downhole formations in the drilling platform control center; According to the method for extracting physical and mechanical property parameters of underground formations, the format consistent integrity data and the adaptation statistical table, periodic physical and mechanical property parameter feature extraction of underground formations at different mining depths is performed, and abnormal features of physical and mechanical property parameters of underground formations at different mining depths are generated; The big data edge computing gateway analyzes the execution status of the soft rock creep test, judges and analyzes the soft rock creep test work for discovering the physical and mechanical property parameters of the downhole formation, and generates abnormal characteristics of the physical and mechanical property parameters of the downhole formation of the soft rock creep test. The abnormal characteristics of the physical and mechanical property parameters of the downhole formation of the soft rock creep test and the abnormal characteristics of the physical and mechanical property parameters of the downhole formation of the different mining depths form the abnormal characteristics.
9. The soft rock creep testing method based on periodic dynamic load according to claim 7 is characterized in that: According to the abnormal characteristics of the unit cycle, the soft rock creep test work of the underground formation of the drilling platform and the physical and mechanical property parameters of the underground formation at different mining depths are analyzed and clustered, and at the same time, the mechanical vibration frequencies of different equipment under the periodic dynamic load at different mining depths are analyzed to form the analytical clustering parameters of the periodic dynamic load at different mining depths. The steps specifically include: locating the rock formation based on the abnormal characteristics of the physical and mechanical property parameters of the underground formation in the soft rock creep test and the abnormal characteristics of the physical and mechanical property parameters of the underground formation at different mining depths, including stress analysis of the physical and mechanical property parameters of the underground formation, temperature analysis of the physical and mechanical property parameters of the underground formation, and abnormal positioning of the physical and mechanical property parameters of the underground formation; Classify and form the abnormal characteristics of physical and mechanical property parameters of underground formations and the physical and mechanical property parameter information of underground formations in the abnormal characteristics of physical and mechanical property parameters of underground formations at different mining depths, and obtain the data processing flow corresponding to the physical and mechanical property parameter information of underground formations, so as to form the reference benchmark of physical and mechanical property parameters of underground formations, the coefficients to be optimized of the calculation model of physical and mechanical property parameters of underground formations, and the types of noise; detect the mechanical vibration of the entire periodic dynamic load at different mining depths, and statistically analyze the sources of periodic dynamic loads of equipment at different mining depths to form analytical clustering parameters of periodic dynamic loads at different mining depths; Modeling different mining depths in the underground formation of the drilling platform to form a soft rock creep model, wherein the soft rock creep model includes a soft rock creep viscoelastic plastic model, a soft rock damage creep model and a soft rock creep numerical simulation model; According to the soft rock creep model and the format consistent integrity data, the corresponding viscosity, elasticity, plastic deformation, damage variables, and creep behavior are tested for soft rock creep, and the influencing factors of soft rock creep are divided, and the steps of completing the soft rock creep safety test of the underground formation of the drilling platform specifically include: constructing and managing the soft rock creep viscoelastic-plastic model, the soft rock damage creep model, and the soft rock creep numerical simulation model; According to the test data at different temperatures, the soft rock creep viscoelastic-plastic model, the soft rock damage creep model and the soft rock creep numerical simulation model, the preset attention mechanism and identification algorithm are used to identify the soft rock creep safety of different mining depths, systems and ground stations to obtain soft rock creep identification parameters; according to the soft rock creep safety identification parameters and the periodic dynamic load source of equipment mining difficulty, the mining difficulty of underground formations at different mining depths is quantitatively evaluated and the mining difficulty identification parameters are obtained; Regularly or according to the soft rock creep identification parameters and mining difficulty identification parameters, the soft rock creep safety identification of different mining depths, systems and ground stations is automatically completed, and a record of influencing factors of soft rock creep test is formed; the soft rock creep process of the system is visualized in 3D.
10. The soft rock creep testing method based on periodic dynamic load according to claim 7, characterized in that: The soft rock creep safety of the underground formation of the drilling platform is modeled and identified by the Burgers algorithm through the soft rock creep safety test parameters of the underground formation of the drilling platform, and the parameters of the Burgers algorithm are updated by using the multi-layer attention mechanism and convolutional neural network optimization. The specific steps include: continuously testing the working status of different mining depths and the periodic dynamic load sources of the mining difficulty of the system equipment, forming a record of different mining depths and the mining difficulty of the drilling platform, and providing a reference for subsequent mining and soft rock creep testing; Determine the soft rock creep safety at different mining depths to obtain the soft rock creep safety test parameters, evaluate the duration of different mining depths or soft rock creep grades, and evaluate the test accuracy of future soft rock creep tests at different mining depths; based on the soft rock creep safety at different mining depths, combined with the test accuracy of future soft rock creep tests at different mining depths, use the multi-layer attention mechanism and convolutional neural network optimization to update the parameters of the Burgers algorithm.
Citation Information
Patent Citations
Low-temperature rock creep test device and system
CN112945750A
Soft rock compression creep similar environment simulation test device and test method
CN113075051A
Method for constructing rock creep damage constitutive model under action of freeze-thaw cycle
CN113742915A
Method, system and equipment for predicting minimum horizontal principal stress profile model of oil and gas reservoir
CN114622907A
Method for constructing non-linear damage creep model of carbonaceous slate with different bedding dip angles
CN117390880A
Cited By
Soft rock rheological mechanical property testing method based on vibration-seepage pressure coupling
CN121364139A
Method for testing rheological mechanical properties of soft rock based on vibration-seepage pressure coupling
CN121364139B
