Ground stress measurement method based on indentation technology and machine learning

Abstract

This invention belongs to the field of geostress measurement and discloses a geostress measurement method based on indentation technology and machine learning. First, core samples are drilled and processed into specimens. Shallow indentation tests are conducted to obtain indentation load-displacement curves and an equivalent elastic model of the core. Deep indentation tests are then conducted, and the test data are used as actual training samples for machine learning. The actual training sample training set is input into a neural network for network training, resulting in a geostress inverse problem model. In-situ indentation tests are performed, and the minimum and maximum horizontal principal stresses are calculated based on the equivalent elastic model of the core. The direction of the principal crack in the indentation at the bottom of the well is measured using an imaging logging instrument, which indicates the direction of the maximum horizontal principal stress. This invention constructs a geostress measurement method for deep Earth rocks based on indentation technology and machine learning. Compared with current conventional rock mechanical property measurement techniques, indentation technology is convenient to operate and can be used for in-situ testing of the mechanical behavior of rocks in deep Earth and soft rock areas with large deformation.

Description

Technical Field

[0001] This invention belongs to the field of geostress measurement, specifically relating to a geostress measurement method based on indentation technology and machine learning. Background Technology

[0002] With increasing human demand for energy and mineral resources and the continuous intensification of mining, shallow mineral resources are becoming increasingly scarce. The proportion of roadways in deep mines with fractured and weak surrounding rock under high ground stress conditions has increased dramatically, and large deformation disasters in soft rock occur frequently. These disasters typically exhibit characteristics such as high stress, large deformation, strong rheology, low strength, and difficulty in support, and have become a focal point and challenging issue in deep mining rock mechanics research.

[0003] Currently, commonly used geostress measurement methods typically simplify rock masses into mechanical models such as elastic, viscoelastic, porosimetric, or ideal elastoplastic models. This results in each commonly used geostress testing method having its own applicable scope and certain limitations.

[0004] Hydraulic fracturing is based on linear elasticity and makes the following assumptions: the rock mass at the measuring point is a continuous, homogeneous, and isotropic linear elastic body; the rock mass at the measuring point is impermeable; and the borehole must be parallel to one of the principal stresses. While this method can measure the minimum horizontal principal stress relatively well, the error in measuring the maximum horizontal principal stress is large, and it cannot be used to measure the in-situ stress in fractured or plastically deformed rock zones. The ground stress relief method assumes that the rock mass is a continuous, homogeneous, and isotropic linear elastic body. Under loading and unloading, the stress and strain experienced by the rock mass have the same functional relationship. It uses elasticity theory to calculate the magnitude and direction of the in-situ stress on the rock mass element. This method has high requirements for the rock mass being measured and is difficult to use for measuring deep in-situ stress. The aniline strain recovery method assumes the rock is a homogeneous, isotropic viscoelastic material. When the core sample separates from the surrounding rock mass, some parts immediately recover elastically, while others recover slowly over time using a hysteretic-elastic method. The amount of strain recovery in each direction is positively correlated with the previously applied pressure. For rocks with high clay mineral content (such as shale), the shrinkage deformation caused by mineral dehydration during measurement is opposite to the expansion deformation caused by stress release. Therefore, the aniline strain recovery method is not suitable for rock samples with high clay content. The borehole breakout method uses the shape of the borehole wall breakout and rock strength parameters to determine the magnitude of the horizontal principal stress, and estimates the stress value based on the depth and width of the breakout. When there is no breakout in the borehole, relevant geostress information cannot be obtained; if the rock is highly anisotropic or heterogeneous, it will also introduce significant errors in determining the geostress value and orientation. The acoustic emission method is a method for measuring geostress based on elasticity theory and utilizing the Kaiser effect of acoustic emission from rocks. However, since acoustic emission is related to elastic wave propagation, high-strength brittle rocks typically exhibit a significant Kaiser effect, while the Kaiser effect is often insignificant in porous, low-strength, and ductile rock masses. Therefore, acoustic emission is generally not recommended for determining stress in weak, loose, and ductile rock masses, and its application is only permitted within the elastic range of the lithology (within 60% of the lithological failure strength). Focal mechanism analysis, based on linear elasticity theory, uses the discrete focal mechanism of each microseismic event and the slip direction of the rupture surface as input parameters to perform geostress inversion based on the actual cross-sectional slip vector, thus obtaining statistically significant geostress. This method can only determine the stress changes caused by earthquakes in the focal region, the direction of spatial tectonic stress in a large area, and the relative magnitudes of the three principal stresses, but cannot obtain absolute values.Since each geostress testing method has its limitations, in recent years multiple methods have often been combined to measure geostress in a region. Geostress testing methods that combine field measurements, numerical simulations, and mechanical modeling have been developed to adapt to various complex geological conditions and improve data accuracy.

[0005] Deep rocks and shallow soft rocks often exhibit significant nonlinear characteristics. However, current geostress testing techniques based on elasticity theory struggle to accurately measure geostress in deep soft rock areas. There is an urgent need to study the nonlinear stress-strain relationship of rocks under high temperature and high confining pressure, and to establish geostress measurement methods suitable for both deep rocks and shallow soft rocks, providing scientific and technological support for deep earth exploration and underground engineering construction. Summary of the Invention

[0006] To address the aforementioned technical problems, this invention provides a geostress measurement method based on indentation technology and machine learning, thereby resolving the issues in the prior art. The technical solution adopted by this invention is as follows:

[0007] A geostress measurement method based on indentation technology and machine learning includes the following steps:

[0008] Step 1: Drill core samples and record the original core temperature, then store the core samples in a moist environment;

[0009] Step 2: Process the core into samples;

[0010] Step 3: Calculate the pressure of the overlying rock strata above the core;

[0011] Step 4: Heat the core sample to the original core temperature, perform a shallow indentation test on the sample using a conical indenter, obtain the indentation load-displacement curve, and obtain the core equivalent elastic model;

[0012] Step 5: Heat the sample to the original core temperature, set the minimum horizontal stress and the maximum horizontal principal stress, and conduct a deep indentation test on the sample using a conical indenter. Use the test data as actual training samples for machine learning.

[0013] Step 6: Expand the actual training samples into a training set; input the training set into the neural network for network training to obtain the inverse geostress problem model;

[0014] Step 7: Conduct in-situ indentation tests to obtain indentation load-displacement curves. Calculate the curvature of the loading curve, the slope of the maximum indentation depth of the unloading curve, and the ratio of residual work to total work. Input the data into the trained neural network to obtain the dimensionless values ​​of the maximum and minimum horizontal principal stresses. Based on the core equivalent elastic model obtained in Step 4, calculate the minimum and maximum horizontal principal stresses.

[0015] Step 8: Use an imaging logging instrument to measure the direction of the main crack in the bottom hole indentation, which is the direction of the maximum horizontal principal stress.

[0016] The present invention has the following beneficial effects:

[0017] This invention constructs a method for measuring geostress in deep Earth rocks based on indentation technology and machine learning. Compared with existing conventional rock mechanical property measurement techniques, indentation technology is convenient to operate and can be used for in-situ testing of the mechanical behavior of deep Earth rocks. Deep Earth rocks have ductile deformation characteristics, and the Bayesian neural network used can construct a nonlinear mechanical model, which can not only characterize the elastic deformation behavior of deep Earth rocks, but also their plastic deformation. This overcomes the shortcomings of existing geostress measurement methods based on elastic theory and further expands the range of deep Earth geostress measurement. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the indentation load-displacement curve of the present invention;

[0019] Figure 2 This is a schematic diagram of the geostress inverse problem model based on Bayesian neural networks of this invention. Detailed Implementation

[0020] The following will be based on embodiments of the present invention. Figure 1-Figure 2 The technical solutions in the embodiments of the present invention will be clearly and completely described. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Unless otherwise specified, the technical means used in the embodiments are conventional means well known to those skilled in the art.

[0021] A geostress measurement method based on indentation technology and machine learning includes the following steps:

[0022] Step 1: Drill core samples and record the original core temperature, and store the core samples in a moist environment.

[0023] Step 2: Process the core into a sample (5cm×5cm×10cm).

[0024] Step 3: Calculate the pressure of the overlying rock layer on the core according to formula (1).

[0025] p = ρgh r (1)

[0026] In the formula, ρ is the density of the rock, g is the acceleration due to gravity, and h is the acceleration due to gravity. r The thickness of the overlying rock.

[0027] Step 4: Heat the core sample to the original core temperature, and use a conical indenter with a 60-degree angle to conduct a shallow indentation test on the sample (the ratio of the indentation depth h to the average microstructure scale L of the core sample is greater than 5, and the ratio to the side length b of the sample cross section is less than 0.2). Obtain the indentation load-displacement curve, and obtain the equivalent elastic model of the core according to Formulas 2-6.

[0028]

[0029] In the formula: E and υ are Young's modulus and Poisson's ratio of soft rock, respectively;

[0030] E i ,v i These are Young's modulus and Poisson's ratio of the indenter, respectively.

[0031] E * Equivalent Young's modulus;

[0032] The slope of the S-unloading curve;

[0033] β is a constant related to the geometry of the indenter; for a conical indenter, the value is 1.05.

[0034] P u Pressure head unloads pressure;

[0035] h Indentation depth;

[0036] h max Maximum indentation depth;

[0037] A. Contact area between the indenter and the core sample;

[0038] h c The contact depth between the indenter and the lithological sample.

[0039] Step 5: Heat the sample to the original core temperature, and set the minimum horizontal stress as ρgh. r ,1.2ρgh r ,1.6ρgh r ,2ρgh r ,2.4ρgh r The maximum horizontal principal stresses are 0.8ρgh. r ,1.44ρgh r 2.56ρgh r ,4ρgh r 5.76ρgh r A conical indenter with a 60-degree apex angle was used to conduct deep indentation tests on the specimens. The ratio of the indentation depth h to the specimen cross-sectional side length b was greater than 0.2 and less than 0.5, resulting in a series of indentation load-displacement curves. The curvature C of the loading curve and the slope of the maximum indentation depth of the unloading curve were calculated for each load-displacement curve. and the ratio of residual work to total work Equivalent eigenvalues ​​can be referenced. Figure 1 , as actual training samples for machine learning.

[0040] Specifically, the curvature C of the loading curve of the core sample indentation load-displacement curve and the slope of the maximum indentation depth of the unloading curve. and the ratio of residual work to total work The relationship between eigenvalues ​​and core mechanical parameters is as follows:

[0041]

[0042] Step Six: As Figure 2 A geostress inverse problem model is constructed based on a Bayesian neural network. A generative adversarial network is used to expand the actual data training set. The training set is then input into the neural network for training. By minimizing the KL divergence of the variational distribution and the posterior distribution of the Bayesian neural network, the network weights are learned, resulting in the geostress inverse problem model.

[0043] Step 7: Conduct an in-situ indentation test to obtain the indentation load-displacement curve, and calculate the curvature C0 of the loading curve and the slope of the maximum indentation depth of the unloading curve. and the ratio of residual work to total work Equivalent feature values ​​are input into a trained neural network to obtain the dimensionless value of the maximum horizontal principal stress. and the dimensionless value of the minimum horizontal principal stress Based on the equivalent elasticity model E obtained in step four * The minimum horizontal principal stress σ is calculated. h and maximum horizontal principal stress σ H .

[0044] Step 8: Use imaging logging instruments to measure the direction of the main crack in the bottom hole indentation, which is the direction of the maximum horizontal principal stress.

[0045] Borehole ultrasonic imaging boasts high resolution, high precision, and the ability to produce clear images even in turbid well fluids, providing a wealth of useful information for fundamental engineering analyses, thus gaining widespread application. For example, the HIRAT (High Resolution Acoustic Televiewer) borehole imaging system from Robertson Geologging in the UK. High-resolution acoustic television is specifically designed to provide qualitative images of the borehole wall. Because it uses ultrasound instead of visible light, it can operate in turbid well water, expanding its application range. Furthermore, its high resolution and accurate orientation determination make it particularly suitable for determining the dip direction, angle, and detailed development of rock fractures, analyzing geological structures, and providing reliable data for geological, hydrogeological research, or major engineering evaluations. The main components of a well logging system include... Figure 1 The Micrologger2 is currently the most powerful portable logging system on the market, equipped with a USB interface for connecting to a laptop, dual DSP (Digital Signal Processing) processors, and a built-in borehole video support module. Lightweight and smaller than a typical laptop, it boasts powerful functionality, supporting all RG probes and cameras, including the latest acoustic and optical imaging systems. The Micrologger2 requires only a PC, a probe, and an RG or third-party supplied winch to provide high-quality logging and drilling television data in any situation.

[0046] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Any modifications, alterations, substitutions, or variations made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention shall fall within the protection scope defined by the claims of the present invention.

Claims

1. A geostress measurement method based on indentation technology and machine learning, characterized in that, Includes the following steps: Step 1: Drill core samples and record the original core temperature, then store the core samples in a moist environment; Step 2: Process the core into samples; Step 3: Calculate the pressure of the overlying rock strata above the core; Step 4: Heat the core sample to the original core temperature, perform a shallow indentation test on the sample using a conical indenter, obtain the indentation load-displacement curve, and obtain the core equivalent elastic model; Step 5: Heat the sample to the original core temperature, set the minimum horizontal stress and the maximum horizontal principal stress, and conduct a deep indentation test on the sample using a conical indenter. Use the test data as actual training samples for machine learning. Step 6: Construct a neural network training set based on actual experimental data; input the training set into the neural network for network training to obtain the inverse geostress problem model; Step 7: Conduct in-situ indentation tests to obtain indentation load-displacement curves. Calculate the curvature of the loading curve, the slope of the maximum indentation depth of the unloading curve, and the ratio of residual work to total work. Input the data into the trained neural network to obtain the dimensionless values ​​of the maximum and minimum horizontal principal stresses. Based on the core equivalent elastic model obtained in Step 4, calculate the minimum and maximum horizontal principal stresses. Step 8: Use an imaging logging instrument to measure the direction of the main crack in the bottom hole indentation, which is the direction of the maximum horizontal principal stress; In step five, a series of indentation load-displacement curves are obtained through deep indentation tests; the curvature of the displacement curve, the slope of the maximum indentation depth of the unloading curve, and the ratio of residual work to total work are calculated for each load, and used as actual training samples for machine learning. Core specimen indentation load-displacement curve loading curve curvature Slope of maximum indentation depth on unloading curve and the ratio of residual work to total work The relationship with core mechanical parameters is as follows: ； ； ； In the formula, It is the equivalent Young's modulus.

2. The geostress measurement method based on indentation technology and machine learning according to claim 1, characterized in that, In step three, the pressure of the overlying rock strata above the core is calculated according to formula (1); ； In the formula, For rock density, It is the acceleration due to gravity. The thickness of the overlying rock.

3. The geostress measurement method based on indentation technology and machine learning according to claim 1, characterized in that, In step four, the equivalent elastic model of the core is obtained using the following formula; ； ； ； ； ； In the formula: These represent Young's modulus and Poisson's ratio for soft rock, respectively. These are Young's modulus and Poisson's ratio of the indenter, respectively; Equivalent Young's modulus; The slope of the unloading curve; These are constants related to the geometry of the indenter; To unload the pressure from the pressure head; Indentation depth; Maximum indentation depth; This represents the contact area between the indenter and the core sample. This represents the contact depth between the indenter and the lithological sample.

4. The geostress measurement method based on indentation technology and machine learning according to claim 1, characterized in that, Step six includes: constructing an inverse geostress problem model based on a Bayesian neural network; expanding the actual training samples into a training set using a generative adversarial network; inputting the training set into the neural network for network training; learning the network weights of the Bayesian neural network by minimizing the KL divergence of the variational distribution and the KL divergence of the posterior distribution of the Bayesian neural network, and obtaining the inverse geostress problem model.

5. The geostress measurement method based on indentation technology and machine learning according to claim 1, characterized in that, When performing the shallow indentation test in step four, the indentation depth The ratio of the average microstructure scale L of the core sample to the sample cross-sectional side length is greater than 5. The ratio is less than 0.

2.

6. The geostress measurement method based on indentation technology and machine learning according to claim 1, characterized in that, In step five, the minimum horizontal stresses are set as follows: ； The maximum horizontal principal stresses are set as follows: ； in For rock density, It is the acceleration due to gravity. The thickness of the overlying rock.

Images