Rock shear failure whole process prediction method based on acoustic emission monitoring

Abstract

The invention relates to a rock shear failure whole process prediction method based on acoustic emission monitoring, which comprises the following steps: firstly, collecting a rock sample, processing the rock sample into a plurality of rock test pieces, manufacturing an acoustic emission mold, secondly, installing an acoustic emission test platform, carrying out multiple acoustic emission tests of the rock shear failure whole process by using the acoustic emission test platform, and recording data of the multiple acoustic emission tests, the method comprises the following steps: selecting and calculating calculation parameters of a rock shear failure whole-process prediction model based on experimental data, finally constructing the rock shear failure whole-process prediction model based on an XGBoost algorithm, inputting the calculation parameters of the rock shear failure whole-process prediction method into the model, and outputting a rock shear failure prediction result. According to the method, the stability of the rock mass can be continuously monitored in real time, different stages in the direct shear process of the rock are predicted in a classified mode, and the safety and reliability of rock mass engineering such as tunnel engineering can be improved.

Description

Technical Field

[0001] The invention belongs to the field of mining, and specifically discloses a method for predicting the whole process of rock shear failure based on acoustic emission monitoring.

[0002] Background

[0003] In the process of tunnel construction in rock mass engineering such as underground tunnel engineering and slope engineering, adverse geological conditions such as fault fracture zones are often encountered. In order to ensure that the tunnel can be excavated safely and smoothly, to ensure the normal development of transportation, and to protect the lives and property of workers, it is crucial to accurately predict the stability of tunnel faults. Traditional rock stability monitoring methods mainly include on-site observation, displacement monitoring, stress monitoring, etc. On-site observation relies heavily on manual experience. Construction workers judge the stability of the rock mass by observing some characteristics of the rock surface with the naked eye, such as the appearance of cracks and the deformation of rocks. However, this method has obvious limitations, because the structure and changes inside the rock mass cannot be directly observed, and it is difficult to accurately predict the potential damage inside the rock mass. Displacement monitoring and stress monitoring usually require the installation of acoustic emission probes on the surface or inside the rock. Displacement monitoring mainly evaluates the stability of the rock by measuring the displacement changes of the rock in different directions, and stress monitoring detects the stress distribution inside the rock. However, these methods can often only obtain local information and cannot fully understand the overall condition of the rock mass. Moreover, for the initiation and expansion of small cracks inside the rock mass, the effect of these monitoring methods is not ideal. In addition, traditional monitoring methods are usually intermittent and cannot monitor rock stability in real time and continuously. In actual engineering, rock stability is a dynamic process. Intermittent monitoring may miss early signs of rock shear failure and make it difficult to capture key information in a timely manner. Under complex geological conditions such as tunnel faults, the limitations of traditional monitoring methods are more prominent and cannot meet the requirements of engineering practice for the accuracy and timeliness of rock stability prediction.

[0004] Based on this, the present invention applies acoustic emission technology to rock engineering. By deeply processing and analyzing the acoustic emission parameter data and using the XGBoost algorithm model in machine learning, the collected acoustic emission parameter data and the rock direct shear test stage are classified and predicted. Finally, a method for predicting the entire process of rock shear failure based on acoustic emission monitoring is proposed, which meets the requirements of engineering practice for accuracy and timeliness of rock stability prediction. Summary of the invention

[0005] In order to solve the problem that existing rock mass monitoring methods are difficult to monitor rock mass stability in real time and continuously, miss early signs of rock shear failure, and have difficulty in capturing key information in time, the present invention proposes a method for predicting the whole process of rock shear failure based on acoustic emission monitoring.

[0006] The present invention provides a method for predicting the whole process of rock shear failure based on acoustic emission monitoring, which is characterized by comprising the following steps:

[0007] S1. Collect rock samples and process them into multiple rock specimens, make an acoustic emission mold according to the size of the rock specimens, and open square holes on the acoustic emission mold;

[0008] S2. Install an acoustic emission test platform, the acoustic emission test platform comprising: a mechanical loading device, an acoustic emission monitoring system and the acoustic emission mold produced in step S1;

[0009] S3. Conduct multiple acoustic emission tests of the whole process of rock shear failure using the acoustic emission test platform installed in step S2, place the rock specimen processed in step S1 into the acoustic emission mold, arrange an acoustic emission probe in the acoustic emission mold, set the parameters of the acoustic emission monitoring system, and record the data of multiple acoustic emission tests;

[0010] S4. Select three acoustic emission parameters with low correlation as calculation parameters of the rock shear failure whole process prediction model, and calculate the calculation parameters of the rock shear failure whole process prediction model based on the acoustic emission test data obtained in step S3;

[0011] S5. Determine the input features and prediction target variables of the XGBoost algorithm, select data from the acoustic emission test data obtained in step S3, calculate the calculation parameters of the rock shear failure whole process prediction model selected in step S4, construct a data set for training the XGBoost algorithm, and train the XGBoost algorithm;

[0012] S6. Initialize the XGBoost algorithm trained in step S5, determine the basic operation mode and operation parameters of the algorithm, and build a prediction model for the whole process of rock shear failure;

[0013] S7. Input the calculation parameters of the rock shear failure whole process prediction method obtained in step S4 into the rock shear failure whole process prediction model constructed in step S6, and output the rock shear failure prediction result.

[0014] According to a method for predicting the entire process of rock shear failure based on acoustic emission monitoring in some embodiments of the present application, the rock specimen in step S1 is a cube specimen, the surface of the rock specimen is smooth, and the error of non-parallelism and non-perpendicularity of the ends is ±0.02 mm;

[0015] The acoustic emission mold also includes a wedge-shaped plug for fixing the acoustic emission probe.

[0016] According to a method for predicting the entire process of rock shear failure based on acoustic emission monitoring in some embodiments of the present application, the mechanical loading device in step S2 is arranged outside the acoustic emission mold to apply axial pressure and shear stress to the rock specimen;

[0017] The acoustic emission monitoring system is arranged outside the acoustic emission mold, and includes an acoustic emission probe, a preamplifier, an acoustic emission acquisition card and acoustic emission processing software running on an electronic device, and is used to collect the acoustic emission transient waveform of the rock specimen in real time;

[0018] The acoustic emission mold is arranged between the rock specimen and the mechanical loading device, and is used to carry the rock specimen and fix the acoustic emission probe;

[0019] The installation spacing of the acoustic emission probes is 10-20 mm.

[0020] According to a method for predicting the entire process of rock shear failure based on acoustic emission monitoring in some embodiments of the present application, step S3 includes:

[0021] Step 301. Place the rock specimen processed in step S1 into the acoustic emission mold, place the acoustic emission probe in contact with the rock specimen through the square hole of the acoustic emission mold, and insert the wedge-shaped plug into the square hole to fix the acoustic emission probe;

[0022] Step 302. Set the acoustic emission monitoring system parameters, and perform mechanical preloading on the rock sample before the mechanical loading device performs formal mechanical loading. After the mechanical preloading is completed, the vertical force data, vertical displacement data, lateral force data, and lateral displacement data are reset to zero;

[0023] Step 303. Load the normal force applied by the mechanical loading device to the initial normal stress value, and then use the mechanical loading device to perform mechanical loading at a shear rate of 0.2 mm / min. During the mechanical loading process, the mechanical loading process is kept synchronous with the data monitoring of the acoustic emission monitoring system. The initial normal stress value is set according to the particle size and normal stress conditions of the rock specimen;

[0024] Step 304. When the mechanical loading in step 303 reaches a peak value, that is, the destruction phase of the acoustic emission test rock specimen ends, the mechanical loading of the mechanical loading device and the acoustic emission test data monitoring of the acoustic emission monitoring system are stopped at the same time, and the acoustic emission test data obtained by monitoring is collected and saved;

[0025] Step 305: Take out the acoustic emission probe and the destroyed rock specimen, and clean the residue in the acoustic emission mold.

[0026] According to a method for predicting the entire process of rock shear failure based on acoustic emission monitoring in some embodiments of the present application, the acoustic emission probes in step S301 are evenly arranged at a position 5 cm above and below the rock specimen;

[0027] The contact portion between the rock specimen and the acoustic emission probe is evenly coated with vaseline with a thickness of 0.1-0.2 mm.

[0028] According to a method for predicting the whole process of rock shear failure based on acoustic emission monitoring in some embodiments of the present application, the acoustic emission monitoring system parameters in step S302 include: number of channels, threshold value, amplification gain and sampling rate, the number of channels is set to 8, the threshold value is set to 45dB, the amplification gain is set to 40dB, and the sampling rate is set to 1MSPS;

[0029] The preloading force of the mechanical preloading is 10%-20% of the expected maximum bearing capacity of the rock specimen.

[0030] According to a method for predicting the whole process of rock shear failure based on acoustic emission monitoring in some embodiments of the present application, the calculation parameters of the whole process prediction model of rock shear failure in step S4 include: the logarithm of the accumulated energy of acoustic emission, the b value and the main frequency A;

[0031] The acoustic emission cumulative energy is as shown in formula (1):

[0032]

[0033] In the formula, E comulative It represents the cumulative energy of acoustic emission from time 0 to time T, E t Represents the energy carried by the acoustic emission signal at time t;

[0034] The logarithm of the acoustic emission cumulative energy is lgE comulative ;

[0035] The b value is as shown in formula (2):

[0036]

[0037] In the formula, B represents the b value, m represents the total number of magnitude classes, and M j represents the median magnitude of the jth level, N j represents the number of acoustic emission events of the j-th magnitude, lg N as shown in formula (3):

[0038] lg N=a-bM (3)

[0039] In the formula, a and b are constants, M represents the magnitude as in formula (4), and N represents the total number of earthquakes with magnitude greater than M.

[0040] M=(A dB +40) / 20 (4)

[0041] In the formula, A dB represents the acoustic emission amplitude;

[0042] The main frequency A is in KHz and is extracted from the acoustic emission signal in the acoustic emission test data collected in step S3 by using fast Fourier transform.

[0043] According to a method for predicting the entire process of rock shear failure based on acoustic emission monitoring in some embodiments of the present application, step S5 includes:

[0044] Step 501. Select data from the acoustic emission test data collected in step S3, calculate the logarithm of the acoustic emission cumulative energy, the b value and the main frequency A, and select the rock shear failure stage as the prediction target variable;

[0045] Step 502: Divide the rock shear failure stage into a compaction stage, an elastic stage, an inelastic stage, and a post-peak failure stage, encode the compaction stage as 1, encode the elastic stage as 2, encode the inelastic stage as 3, and encode the post-peak failure as 4;

[0046] Step 503. Perform corresponding rock shear failure stage encoding on the logarithm of the acoustic emission cumulative energy, b value and main frequency A of each rock specimen to form labeled data and construct a data set for training the XGBoost algorithm;

[0047] Step 504: Divide the stress stage of the rock specimen into an elastic stage, a yield stage, a strengthening stage and a failure stage, and use the elastic stage, the yield stage, the strengthening stage and the failure stage as input features of the XGBoost algorithm;

[0048] Step 505: Using the compaction stage, the elastic stage, the inelastic stage or the post-peak failure stage of the rock specimen as the target variable of the XGBoost algorithm;

[0049] Step 506: Train the XGBoost algorithm.

[0050] According to a method for predicting the entire process of rock shear failure based on acoustic emission monitoring in some embodiments of the present application, step S6 includes:

[0051] Step 601. Determine the basic operation mode, set the algorithm to run the data in the tree structure gbtree, build an algorithm similar to the tree structure to gradually analyze the data, and set general parameters;

[0052] Step 602: Set the XGBoost algorithm boost parameters, where:

[0053] The maximum number of iterations n_estimator is set to 100,

[0054] The iteration step learning_rate is set to 0.7,

[0055] The step size of each iteration is set to 0.7.

[0056] The maximum depth of the tree is set to 1.5.

[0057] The subsample ratio of random sampling for each tree is set to 1.

[0058] Control the proportion of randomly sampled columns for each tree. Set colsample_bytree to 1.

[0059] The proportion of column sampling colsample_bylevel is set to 1 for each node split in each tree.

[0060] The weight L1 regularization term reg_alpha is set to 0,

[0061] The weight L2 regularization term reg_lambda is set to 0,

[0062] The minimum leaf node sample weight is set to 1.

[0063] min_child_weight is set to 1;

[0064] Step 603. Set the XGBoost algorithm learning target parameters, where:

[0065] Objective is used to specify the logical relationship between the learning input features and the rock shear failure stage. The input features include the logarithm of the acoustic emission cumulative energy, b value and main frequency A. The XGBoost algorithm selects binary:logistic binary classification logistic regression, and the output is the probability that the rock specimen is in the unbroken, compaction stage, elastic stage, inelastic stage and post-peak failure stage.

[0066] eval_metric is a quantitative evaluation of valid data. The XGBoost algorithm takes "auc", which is the area under the ROC curve, and is set to evaluate the performance of the model by calculating the area under the ROC curve of the receiver operating characteristic curve.

[0067] seed is the random number seed, set to 1.

[0068] According to a method for predicting the entire process of rock shear failure based on acoustic emission monitoring in some embodiments of the present application, the general parameters set in step S601 include: turning on the silent mode silent to 0, outputting algorithm operation information; and setting the multi-thread control parameter nthread to the maximum number of computer threads.

[0069] The technical solution provided by the present invention processes and analyzes the acoustic emission parameter data in depth, studies the relationship between the acoustic emission parameter data and the different stages of rock direct shear, and adopts a method of comprehensive analysis of three acoustic emission parameters with low correlation. These three parameters reflect the changes inside the rock from different angles. By comprehensively considering the three parameters of energy parameter, b value and main frequency, the limitations of single parameter analysis can be avoided, and the change information inside the rock can be captured more comprehensively and accurately. This comprehensive analysis method can improve the accuracy and reliability of rock shear failure prediction, so that the prediction results are more in line with the actual situation. Whether in the rock direct shear test in the laboratory or in the application scenario of actual engineering, it can provide more powerful support for the evaluation and prediction of rock stability. In addition, the technical solution provided by the present invention establishes a classification prediction model for different stages of rock direct shear based on acoustic emission data using the XGBoost algorithm in machine learning. In the process of model establishment, by analyzing and processing a large amount of rock direct shear test data, appropriate characteristic parameters are selected, and the rock shear failure stage is reasonably divided, so that the relevant information of the internal structure change of the rock can be collected in real time and stably. By continuously optimizing the parameter settings of the algorithm, the accuracy of the model reached 77%. This high accuracy indicates that the model has good performance and can effectively classify and predict different stages in the rock direct shear process, meeting the requirements of engineering practice for accuracy and timeliness of rock stability prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 A schematic flow chart of a method for predicting the entire process of rock shear failure based on acoustic emission monitoring provided in Example 1 of the present invention;

[0071] Figure 2 A schematic diagram of an acoustic emission mold and a wedge-shaped plug model provided in Example 3 of the present invention;

[0072] Figure 3 A schematic diagram of the XGBoost algorithm flow provided in Example 3 of the present invention;

[0073] Figure 4 A comparison chart of the number of true values ​​after prediction and the number of correct values ​​after prediction at each stage provided in Example 3 of the present invention;

[0074] Figure 5 This is a diagram showing the predicted effect of a CSY-3 specimen provided in Example 3 of the present invention;

[0075] Figure 6 This is a diagram showing the predicted effect of a ZSY-4 specimen provided in Example 3 of the present invention;

[0076] Figure 7 This is a diagram showing the predicted effect of the XSY-3 specimen provided in Example 3 of the present invention. DETAILED DESCRIPTION

[0077] The following embodiments of the present invention are described in further detail in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0078] Embodiment 1: This embodiment provides a method for predicting the whole process of rock shear failure based on acoustic emission monitoring, such as Figure 1 As shown, the following steps are included:

[0079] S1. Collect rock samples and process them into multiple rock specimens, make an acoustic emission mold according to the size of the rock specimens, and open square holes on the acoustic emission mold.

[0080] Preferably, in this embodiment, according to the engineering situation, yellow sandstone rocks taken from Chuxiong, Yunnan are selected as samples. The rocks in this area are mainly yellow sandstone, and samples with different degrees of weathering are included to better simulate the rock diversity in actual engineering. The rock specimen is a cube specimen with a size of 100*100*100. The rock specimen has good integrity, a smooth surface, no obvious defects, and the end non-parallelism and non-verticality errors are ±0.02mm. The size and position of the square hole are determined according to the size and monitoring requirements of the acoustic emission probe. The spacing between the acoustic emission probes is 10-20mm. The acoustic emission mold also includes a wedge-shaped plug for fixing the acoustic emission probe. The size and shape of the wedge-shaped plug match the square hole and the acoustic emission probe to ensure that the probe can be stably fixed during the test and achieve the goal of acoustic emission monitoring during the test.

[0081] S2. Install an acoustic emission test platform, which includes: a mechanical loading device, an acoustic emission monitoring system, and the acoustic emission mold produced in step S1. Preferably, in this embodiment, the mechanical loading device is a rock direct shear rheological testing machine produced by Jilin Guanteng Automation Technology Co., Ltd., a GTC550 full digital controller and rheological test software. The AE monitoring equipment adopts the PCI-2 multi-channel AE monitoring system produced by PAC, a physical acoustics company in the United States. In this embodiment, a mechanical loading device is used to provide axial pressure and shear stress to the rock, the acoustic emission transient waveform of the rock is collected in real time through the acoustic emission monitoring system, and the acoustic emission probe is fixed with an improved acoustic emission mold to realize acoustic emission monitoring of the rock specimen during the test.

[0082] S3. Use the acoustic emission test platform installed in step S2 to conduct multiple acoustic emission tests on the entire process of rock shear failure, place the rock specimen processed in step S1 into the acoustic emission mold, arrange an acoustic emission probe in the acoustic emission mold, set the acoustic emission monitoring system parameters, and record data from multiple acoustic emission tests.

[0083] S4. Select three acoustic emission parameters with low correlation as calculation parameters of the rock shear failure whole process prediction model, and calculate the calculation parameters of the rock shear failure whole process prediction model based on the acoustic emission test data obtained in step S3.

[0084] S5. Determine the input features and prediction target variables of the XGBoost algorithm, select data from the acoustic emission test data obtained in step S3, calculate the calculation parameters of the rock shear failure full process prediction model selected in step S4, construct a data set for training the XGBoost algorithm, and train the XGBoost algorithm.

[0085] S6. Initialize the XGBoost algorithm trained in step S5, determine the basic operation mode and operation parameters of the algorithm, and build a prediction model for the entire rock shear failure process.

[0086] S7. Input the acoustic emission parameters selected in step S4 into the rock shear failure whole process prediction model constructed in step S6, and output the rock shear failure prediction result.

[0087] The whole process prediction method of rock shear failure based on acoustic emission monitoring provided in this embodiment makes full use of the phenomenon that when microcracks are generated, expanded and fractured inside the rock during stress deformation, instantaneous strain energy is released in the form of elastic stress waves. By installing acoustic emission probes on the surface or inside the rock, these acoustic emission signals are monitored in real time, and relevant information about changes in the internal structure of the rock is obtained. The real-time, non-destructive and comprehensive nature of acoustic emission technology is utilized to study the relationship between acoustic emission parameter data and different stages of rock direct shear, and the XGBoost algorithm model in machine learning is used to classify and predict the collected acoustic emission parameter data and rock direct shear test stages. Finally, a whole process prediction method of rock shear failure based on acoustic emission monitoring is proposed, which can collect relevant information about changes in the internal structure of the rock in real time and stably, and meet the requirements of engineering practice for accuracy and timeliness of rock stability prediction.

[0088] In addition, this embodiment adopts a method of comprehensive analysis of three acoustic emission parameters with low correlation. These three parameters reflect the changes inside the rock from different angles. By comprehensively considering the three parameters of energy, b value, and main frequency, the limitations of single parameter analysis can be avoided, and the change information inside the rock can be captured more comprehensively and accurately. This comprehensive analysis method can improve the accuracy and reliability of rock shear failure prediction, making the prediction results more in line with actual conditions. Whether in the direct shear test of rock in the laboratory or in the application scenarios of actual engineering, it can provide more powerful support for the evaluation and prediction of rock stability.

[0089] Embodiment 2: This embodiment provides a method for predicting the entire process of rock shear failure based on acoustic emission monitoring, comprising the following steps:

[0090] S1. Collect rock samples and process them into multiple rock specimens, make an acoustic emission mold according to the size of the rock specimens, and open square holes on the acoustic emission mold.

[0091] S2. Install an acoustic emission test platform, which includes: a mechanical loading device, an acoustic emission monitoring system and the acoustic emission mold produced in step S1. Preferably, in this embodiment, the mechanical loading device is arranged outside the acoustic emission mold, and is used to apply axial pressure and shear stress to the rock specimen. The acoustic emission monitoring system is arranged outside the acoustic emission mold, and includes an acoustic emission probe, a preamplifier, an acoustic emission acquisition card and an acoustic emission processing software running on an electronic device, which is used to collect the acoustic emission transient waveform of the rock specimen in real time. The acoustic emission mold is arranged between the rock specimen and the mechanical loading device, and is used to carry the rock specimen and fix the acoustic emission probe. The installation spacing of the acoustic emission probes is 10-20mm.

[0092] Step 301. Place the rock specimen processed in step S1 into the acoustic emission mold, place the acoustic emission probe in contact with the rock specimen through the square hole of the acoustic emission mold, and insert the wedge-shaped plug into the square hole to fix the acoustic emission probe. Preferably, in this embodiment, the acoustic emission probes are evenly arranged at a position 5 cm above and below the rock specimen, and the contact part between the rock specimen and the acoustic emission probe is evenly coated with vaseline with a thickness of 0.1-0.2 mm.

[0093] Step 302. Set the acoustic emission monitoring system parameters, and perform mechanical preloading on the rock specimen before the mechanical loading device performs formal mechanical loading. After the mechanical preloading is completed, the vertical force data, vertical displacement data, lateral force data, and lateral displacement data are reset to zero. Preferably, the acoustic emission monitoring system parameters in this embodiment include: number of channels, threshold value, amplification gain, and sampling rate. The number of channels is set to 8, the threshold value is set to 45dB, the amplification gain is set to 40dB, and the sampling rate is set to 1MSPS. The preloading force of the mechanical preloading is 10%-20% of the expected maximum bearing capacity of the rock specimen.

[0094] Step 303. Load the normal force applied by the mechanical loading device to the initial normal stress value, and then use the mechanical loading device to perform mechanical loading at a shear rate of 0.2 mm / min. During the mechanical loading process, keep the mechanical loading process synchronized with the data monitoring of the acoustic emission monitoring system. The initial normal stress value is set according to the particle size and normal stress conditions of the rock specimen.

[0095] Step 304. When the mechanical loading of step 303 reaches a peak value, that is, the destruction phase of the acoustic emission test rock specimen ends, the mechanical loading of the mechanical loading device and the acoustic emission test data monitoring of the acoustic emission monitoring system are stopped simultaneously, and the acoustic emission test data obtained by monitoring are collected and saved.

[0096] Step 305: Take out the acoustic emission probe and the destroyed rock specimen, clean the residue in the acoustic emission mold, and prepare for the next test.

[0097] S4. Select three acoustic emission parameters with low correlation as calculation parameters of the rock shear failure whole process prediction model, and calculate the calculation parameters of the rock shear failure whole process prediction model based on the acoustic emission test data obtained in step S3.

[0098] Preferably, the calculation parameters of the rock shear failure whole process prediction model in step S4 of this embodiment include: the logarithm of the acoustic emission cumulative energy, the b value and the main frequency A;

[0099] The acoustic emission cumulative energy is as shown in formula (1):

[0100]

[0101] In the formula, E comulative It represents the cumulative energy of acoustic emission from time 0 to time T, E t Represents the energy carried by the acoustic emission signal at time t, then the logarithm of the accumulated energy of the acoustic emission is lgE comulative .

[0102] The b value is as shown in formula (2):

[0103]

[0104] In the formula, B represents the b value, m represents the total number of magnitude classes, and M j represents the median magnitude of the jth level, N j represents the number of acoustic emission events of the j-th magnitude, lg N as shown in formula (3):

[0105] lg N=a-bM (3)

[0106] In the formula, a and b are constants, M represents the magnitude as in formula (4), and N represents the total number of earthquakes with magnitude greater than M.

[0107] M=(A dB +40) / 20 (4)

[0108] In the formula, A dB represents the acoustic emission amplitude;

[0109] The main frequency A is in KHz and is extracted from the acoustic emission signal in the acoustic emission test data collected in step S3 using fast Fourier transform. In this embodiment, the acoustic emission signal in the acoustic emission test data is transformed into a frequency domain waveform by fast Fourier transform, and the frequency corresponding to the maximum amplitude point in the frequency domain waveform is the main frequency of the acoustic emission.

[0110] Step 501. Select data from the acoustic emission test data collected in step S3, calculate the logarithm of the acoustic emission cumulative energy, the b value and the main frequency A, and select the rock shear failure stage as the prediction target variable. Preferably, in this embodiment, whether the rock specimen is broken or in a certain stage of breaking is used as the target variable.

[0111] Step 502. Divide the rock shear failure stage into a compaction stage, an elastic stage, an inelastic stage and a post-peak failure stage, encode the compaction stage as 1, encode the elastic stage as 2, encode the inelastic stage as 3, and encode the post-peak failure as 4.

[0112] Step 503. Perform corresponding rock shear failure stage encoding on the logarithm of the acoustic emission cumulative energy, b value and main frequency A of each rock specimen to form labeled data and construct a data set for training the XGBoost algorithm.

[0113] Step 504. Divide the stress stage of the rock specimen into an elastic stage, a yield stage, a strengthening stage and a failure stage, and use the elastic stage, the yield stage, the strengthening stage and the failure stage as input features of the XGBoost algorithm.

[0114] Step 505. The compaction stage, the elastic stage, the inelastic stage or the post-peak failure stage of the rock specimen is used as the target variable of the XGBoost algorithm.

[0115] Step 506: Train the XGBoost algorithm.

[0116] Step 601. Determine the basic operation mode, set the algorithm to the tree structure gbtree to run the data, build an algorithm similar to the tree structure to gradually analyze the data, and set general parameters. Preferably, the general parameters in this embodiment include: turning on the silent mode silent to 0, outputting the algorithm operation information; and setting the multi-thread control parameter nthread to the maximum number of threads on the computer.

[0117] Step 602: Set the XGBoost algorithm boost parameters, where:

[0118] The maximum number of iterations n_estimator is set to 100,

[0119] The iteration step learning_rate is set to 0.7,

[0120] The step size of each iteration is set to 0.7.

[0121] The maximum depth of the tree is set to 1.5.

[0122] The subsample ratio of random sampling for each tree is set to 1.

[0123] Control the proportion of randomly sampled columns for each tree. Set colsample_bytree to 1.

[0124] The proportion of column sampling colsample_bylevel is set to 1 for each node split in each tree.

[0125] The weight L1 regularization term reg_alpha is set to 0,

[0126] The weight L2 regularization term reg_lambda is set to 0,

[0127] The minimum leaf node sample weight is set to 1.

[0128] min_child_weight is set to 1.

[0129] Step 603. Set the XGBoost algorithm learning target parameters, where:

[0130] Objective is used to specify the logical relationship between the learning input features and the rock shear failure stage. The input features include the logarithm of the acoustic emission cumulative energy, b value and main frequency A. The XGBoost algorithm selects binary:logistic binary classification logistic regression, and the output is the probability that the rock specimen is in the compaction stage, the elastic stage, the inelastic stage and the post-peak failure stage.

[0131] eval_metric is a quantitative evaluation of valid data. The XGBoost algorithm takes "auc", which is the area under the ROC curve, and is set to evaluate the performance of the model by calculating the area under the ROC curve of the receiver operating characteristic curve.

[0132] seed is the random number seed, set to 1.

[0133] S7. Input the calculation parameters of the rock shear failure whole process prediction method obtained in step S4 into the rock shear failure whole process prediction model constructed in step S6, and output the rock shear failure prediction result.

[0134] Example 3. This example provides a method for predicting the entire process of rock shear failure based on acoustic emission monitoring. Taking the example of establishing a rock direct shear process classification prediction model based on machine learning for yellow sandstone of different particle sizes, the method for predicting the entire process of rock shear failure based on acoustic emission monitoring is explained in detail.

[0135] The design background of the whole process prediction method of rock shear failure based on acoustic emission monitoring in this embodiment: In the excavation and construction process of many rock projects, mining projects, underground tunnel projects, slope projects and other projects often face complex geological conditions. Among them, faults are a common unfavorable geological structure, and the mechanical properties of the rock mass are poor. There are a large number of cracks and weak surfaces inside the rock, which makes the rock mass very easy to shear failure when subjected to external stress. This shear failure may cause deformation, collapse and other disasters of the rock mass, which will not only have a serious impact on the progress of the project, but also threaten the life safety of the construction workers and the stability of the surrounding environment.

[0136] Yellow sandstone is a common type of rock in geological structures and widely distributed in the strata traversed by many tunnel projects. These physical and mechanical properties of yellow sandstone are representative, and it is widely distributed in the strata involved in tunnel projects. Through detailed indoor test research on yellow sandstone and analysis of its mechanical behavior and acoustic emission characteristics under different conditions, we can gain a deep understanding of the mechanism of rock shear failure, and then establish an effective rock stability evaluation and prediction method, providing strong technical support for the prediction of tunnel fault stability.

[0137] Overview of mechanical properties of yellow sandstone: The density of yellow sandstone is usually 2.2-2.6g / cm 3 Between (the average value in a certain area is about 2.4g / cm 3 ), the porosity is generally around 10%-25% (about 12% for fine particles and about 20% for coarse particles). In terms of mechanical properties, the uniaxial compressive strength varies depending on factors such as particle size and degree of cementation. Coarse particles are 20-50MPa, medium particles are 30-60MPa, and fine particles are 40-80MPa. The compressive strength increases with the increase of confining pressure.

[0138] This embodiment provides a method for predicting the entire process of rock shear failure based on acoustic emission monitoring, comprising the following steps:

[0139] Step S1. Collect rock samples and process them into multiple rock specimens, make an acoustic emission mold according to the size of the rock specimens, and open square holes on the acoustic emission mold.

[0140] Preferably, in this embodiment, according to the actual engineering situation, yellow sandstone rocks taken from Chuxiong, Yunnan are selected as samples. The rocks in this area are mainly yellow sandstones, and samples of different weathering degrees are included to better simulate the rock diversity in actual engineering. During the collection process, professional rock collection tools are used to avoid unnecessary damage to the rocks. The collected rock samples are transported to the laboratory, and the yellow sandstone specimens are processed into cube rock specimens with a size of 100*100*100, 14 coarse-grained yellow sandstones, 5 medium-grained and fine-grained yellow sandstones, and are numbered CSY1-14, ZSY1-5, and XSY1-5 respectively. CSY represents coarse sandstone, ZSY represents medium sandstone, and XSY represents fine sandstone. The rock specimens have good integrity, smooth surface, no obvious defects, and the non-parallelism and non-verticality of the ends are controlled within ±0.02mm.

[0141] Preferably, in this embodiment, the Figure 2 The direct shear AE mold shown in the figure has square holes on it. The size and position of the holes are determined according to the size of the AE probe and the monitoring requirements. Reasonably select the spacing between the AE probes. According to the size of the rock specimen and the physical properties of the rock specimen, select a spacing of 10-20mm to ensure the strength and accuracy of the AE signal and avoid mutual interference between the AE probes. Set up a wedge-shaped plug for fixing the AE probe. The size and shape of the wedge-shaped plug match the hole and the probe to ensure that the AE probe can be stably fixed during the test and achieve the goal of AE monitoring during the test.

[0142] S2. Install an acoustic emission test platform, which includes: a mechanical loading device, an acoustic emission monitoring system, and the acoustic emission mold produced in step S1.

[0143] Preferably, the present implementation includes debugging and installation and direct shear acoustic emission molds, etc. The instruments used in this test include: the mechanical loading equipment used is a rock direct shear rheological testing machine produced by Jilin Guanteng Automation Technology Co., Ltd., a GTC550 fully digital controller and rheological test software. The AE monitoring equipment adopts the PCI-2 multi-channel AE monitoring system produced by PAC, a physical acoustics company in the United States. In this embodiment, a mechanical loading device is used to provide axial pressure and shear stress to the rock, the acoustic emission transient waveform of the rock is collected in real time through the acoustic emission monitoring system, and the acoustic emission probe is fixed with an improved acoustic emission mold to realize acoustic emission monitoring of the rock specimen during the test.

[0144] S3. Implementation of the test plan: using the acoustic emission test platform installed in step S2 to conduct multiple acoustic emission tests on the entire rock shear failure process; placing the rock specimen processed in step S1 into the acoustic emission mold; arranging an acoustic emission probe in the acoustic emission mold; setting the acoustic emission monitoring system parameters; and recording multiple acoustic emission test data.

[0145] Preferably, in this embodiment, the processed rock specimen is placed in the installed acoustic emission mold, and according to the pre-designed acoustic emission probe arrangement scheme, 8 acoustic emission probes are placed through the mold holes to contact the specimen, and are evenly arranged at a position 5 cm above and below the specimen. Vaseline is evenly applied to the contact part between the rock specimen and the acoustic emission probe, and the amount of application is preferably sufficient to ensure a good coupling effect, and the general application thickness is about 0.1-0.2 mm. Insert the wedge-shaped plug into the hole and fix the acoustic emission probe. The parameter setting scheme of the sliding window length and the sliding distance in the main frequency data processing is shown in Table 1.

[0146] Table 1 Scheme for setting the sliding window length and sliding distance parameters in main frequency data processing

[0147]

[0148]

[0149] Preferably, in this embodiment, the acoustic emission monitoring system parameters are set, and according to the predetermined parameter settings, the number of channels is set to 8, the threshold value is set to 45dB, the amplification gain is set to 40dB, and the sampling rate is set to 1MSPS. The rock sample is preloaded before loading, and the preload force is determined to be 10%-20% of the maximum force that the sample is expected to withstand according to the size and properties of the sample. For example, for a cube specimen with a side length of 100mm, the maximum force expected to withstand is 100kN, and the preload force is 10-20kN. After the preloading is completed, the vertical force and displacement as well as the lateral force and displacement data are zeroed. Then, the normal force is applied according to the test requirements. The normal force setting value is determined according to the actual engineering situation and the test requirements. In this embodiment, it is set to four cases of 6MPa, 2MPa, 4Mpa and 8MPa. Loading is performed at a shear rate of 0.2mm / min.

[0150] Preferably, in this embodiment, according to the test design, the normal force is applied to the set value. In this embodiment, different initial normal stress values ​​are set for samples with different particle sizes and different normal stress conditions, including four cases: 6MPa, 2MPa, 4Mpa and 8MPa. Then the direct shear tester is loaded at a shear rate of 0.2mm / min. During the loading process, the loading process must be synchronized with the acoustic emission monitoring to ensure that the mechanical response and acoustic emission signals of the rock can be obtained at the same time. When the loading reaches the end of the post-peak destruction stage, the mechanical loading system and the acoustic emission monitoring system are stopped at the same time, and the test data is saved. Afterwards, carefully take out the acoustic emission probe, take out the damaged specimen, carefully clean the rock residue in the mold, and prepare for the next test.

[0151] S4. Select three acoustic emission parameters with low correlation as calculation parameters of the rock shear failure whole process prediction model, and calculate the calculation parameters of the rock shear failure whole process prediction model based on the acoustic emission test data obtained in step S3.

[0152] Preferably, after obtaining the acoustic emission test data in this embodiment, according to the correlation of the acoustic emission parameters, three acoustic emission parameters with low correlation, namely energy parameter, b value and main frequency, are used and calculated as the parameters used in the whole process prediction method of rock shear failure.

[0153] Preferably, in this embodiment, for the acoustic emission energy Es, the energy utilization rate and the acoustic emission cumulative energy are solved according to the acoustic emission energy Es, and then the logarithm of the acoustic emission energy contribution rate and the logarithm of the acoustic emission cumulative energy (i.e., 1g acoustic emission cumulative energy) are taken. The setting scheme of the sliding window length and sliding distance parameters in the main frequency data processing is shown in Table 2.

[0154] Table 2 Sliding window length and sliding distance parameter settings in main frequency data processing

[0155]

[0156] S5. Determine the input features and prediction target variables of the XGBoost algorithm, select data from the acoustic emission test data obtained in step S3, calculate the calculation parameters of the rock shear failure full process prediction model selected in step S4, construct a data set for training the XGBoost algorithm, and train the XGBoost algorithm.

[0157] Preferably, this implementation selects data from the acoustic emission data during the rock direct shear process to establish the data set required for the training algorithm, and selects 1g acoustic emission cumulative energy, b value and main frequency in the acoustic emission parameters, and takes the rock shear failure stage as the classification prediction object. According to the mechanical properties and failure mode during the rock direct shear process, the rock shear failure stage is divided into the compaction stage, the elastic stage, the inelastic stage and the post-peak failure stage. These four stages are encoded respectively, and the codes are 1, 2, 3 and 4.

[0158] Preferably, in this implementation, a detailed correspondence is established for the test data, that is, for each rock sample's three characteristic values ​​of lg acoustic emission cumulative energy, b value and main frequency, the corresponding rock shear failure stage code is clearly defined, thereby forming labeled data. At the same time, these four stress stages (such as elastic stage, yield stage, strengthening stage, failure stage, etc.) are used as input features, and whether the sandstone is broken or in a certain stage of breaking is used as the target variable.

[0159] S6. Initialize the XGBoost algorithm trained in step S5, determine the basic operation mode and operation parameters of the algorithm, and construct the rock shear failure whole process prediction model. Preferably, the general settings of the XGBoost algorithm used in this embodiment are as follows. The algorithm is divided into two parts: model training and test verification. The model is trained iteratively, and new classifiers are added to fit the training errors in the current iteration to optimize the model's fitting effect on the samples. In the test set verification stage, the test set is used to verify the classification prediction performance of the model. The XGBoost algorithm process is as follows: Figure 3 As shown. First determine the basic operation mode, set the algorithm to the tree structure gbtree to run the data, and build an algorithm similar to the tree structure to gradually analyze the data. At the same time, set general parameters, such as whether to turn on the silent mode, silent is 0, that is, output the algorithm operation information, which is convenient for understanding its process; the multi-thread control parameter nthread is set to the maximum number of computer threads to make full use of computer resources to improve the operation speed.

[0160] Preferably, the XGBoost algorithm boost parameters used in the model of this embodiment are as follows:

[0161] (1) The maximum number of iterations n_estimator is set to 100, and 100 iterations are performed to optimize the model.

[0162] (2) The iteration step learning_rate is set to 0.7. The step size of each iteration is set to 0.7. The model prediction accuracy is high and the model runs fast.

[0163] (3) The maximum depth of the tree is set to 1.5 to control the complexity of the model and prevent overfitting. Too deep a tree may over-learn the noise in the training data and affect its performance on new data.

[0164] (4) The subsample ratio of random sampling for each tree, colsample_bytree, which controls the proportion of the number of columns randomly sampled for each tree, and colsample_bylevel, which controls the proportion of columns sampled each time a node is split, are all set to 1 to ensure that the model fully learns the data features and avoids overfitting or underfitting.

[0165] (5) The weighted L1 regularization term reg_alpha is set to 0 to speed up the algorithm; the weighted L2 regularization term reg_lambda is set to 0 to prevent overfitting.

[0166] (6) The minimum leaf node sample weight and min_child_weight are set to 1 to avoid overfitting.

[0167] Preferably, the XGBoost algorithm learning target parameters used in this embodiment are as follows:

[0168] (1) objective: used to specify the learner and the learning goal. The model in this paper selects binary:logistic, two-class logistic regression, and the output is probability. When predicting the rock shear failure stage, the model will learn the logical relationship between the input features (lg acoustic emission cumulative energy, b value and main frequency) and the rock shear failure stage (coded as 1-4), and output the probability of the rock being in a certain stage.

[0169] (2) eval_metric: For the quantitative evaluation of valid data, the model in this paper takes "auc", which is the area under the ROC curve. It is set to evaluate the performance of the model by calculating the area under the receiver operating characteristic curve (ROC) curve.

[0170] (3)seed: random number seed, set to 1 as the random number seed. When the model is trained, a fixed random number seed is set to randomly shuffle the data and initialize some random parameters.

[0171] S7. Input the calculated LG AE cumulative energy, b value and main frequency into the set XGBoost algorithm model to predict the rock shear failure stage and output the prediction results.

[0172] Preferably, this embodiment further includes step S8. ACC evaluation index model evaluation and application.

[0173] In this embodiment, step S8 includes:

[0174] S801. Set indicators and two performance curves to evaluate the effectiveness of the model.

[0175] (1) Accuracy (ACC): It is defined as the ratio of the number of correctly classified samples to the total number of samples, which reflects the classification accuracy of the overall data. However, the ACC index is more affected by the majority class classification data set. For data sets with an unbalanced number of categories, the ACC evaluation index is not comprehensive enough. The ACC evaluation index is shown in formula (5):

[0176]

[0177] Where ACC is the accuracy; TP, TN, FP, and FN are the number of true positive, true negative, false positive, and false negative samples, respectively.

[0178] Whether the sample is correctly classified is determined based on the prediction results of the three characteristics of lg acoustic emission cumulative energy, b value and main frequency A. For example, if the ACC evaluation index model predicts that a rock sample is in failure stage 3 (inelastic stage) and the actual stage is also 3, then the sample is correctly classified and counted in the number of correctly classified samples.

[0179] (2) Balanced Accuracy, BACC: defined as the average accuracy. Since the ACC evaluation index is sometimes not comprehensive enough, BACC is calculated again. The classification accuracy of each rock shear failure stage is calculated separately and the average is taken. That is, the classification accuracy of the compaction stage, elastic stage, inelastic stage and post-peak failure stage is calculated separately, and then the average is obtained to obtain BACC, as shown in formula (6):

[0180]

[0181] Where BACC is the average accuracy; TP, TN, FP, and FN are the numbers of true positive, true negative, false positive, and false negative samples.

[0182] (3) F1-score: It is defined as the arithmetic mean of precision and recall. It combines the results of precision and recall and can comprehensively and objectively reflect the performance of the model. The precision and recall are calculated based on the prediction results of the three features of lg acoustic emission cumulative energy, b value and main frequency, and then the F1-score is calculated, as shown in formula (7):

[0183]

[0184] In the formula, F1-score is the arithmetic mean of precision and recall; Precision is the precision, and Recall is the recall.

[0185] (4) Receiver Operating Characteristic (ROC) curve: It is one of the indicators used to reflect the performance of the classifier. It is a relationship curve between the true positive rate (TPR) and the false positive rate (FPR). The area under the curve (AUC) is a quantitative value for evaluating the performance of the classifier. FPR is shown in formula (8):

[0186]

[0187] In the formula, FPR is the false positive rate of classification, TN is the number of true negative samples, FP is the number of false positive samples, and FN is the number of false negative samples.

[0188] TPR is shown in formula (9):

[0189]

[0190] In the formula, TPR is the true classification rate, TP is the number of true samples, FP is the number of false positive samples, and FN is the number of false negative samples.

[0191] (5) Precision-Recall (PR) curve: It can also reflect the performance of the classifier. It is the relationship curve between precision and recall. Its AUC can be calculated to evaluate the performance of the model.

[0192] Step S802. Use the set indicators and performance curves to evaluate the effect of the rock shear failure prediction model for the whole process, and perform performance analysis on the rock shear failure prediction model for the whole process. After evaluation, the rock direct shear classification prediction model based on acoustic emission data established in this embodiment has an accuracy rate of 77%. Comparison of the number of each stage between the true value and the predicted correct value Figure 4The rock shear failure whole process prediction model was applied to the rock direct shear test of the prepared rock sample. The prediction effect of the CSY-3 specimen is shown in Figure 5 , the prediction effect of ZSY-4 specimen is as follows Figure 6 , the prediction effect of XSY-3 specimen is as follows Figure 7 It can be seen from the figure that the rock shear failure whole process prediction model of this embodiment has a better effect on the classification prediction of the direct shear stage of a single type of rock.

[0193] The embodiments of the present invention are given for the purpose of illustration and description, and are not intended to be exhaustive or to limit the invention to the disclosed forms. Many modifications and variations will be apparent to those of ordinary skill in the art. The embodiments are selected and described in order to better illustrate the principles and practical applications of the present invention and to enable those of ordinary skill in the art to understand the present invention and thereby design various embodiments with various modifications suitable for specific uses.

Claims

1. A method for predicting the whole process of rock shear failure based on acoustic emission monitoring, characterized in that: The steps include: S1. Collect rock samples and process them into multiple rock specimens, make an acoustic emission mold according to the size of the rock specimens, and open square holes on the acoustic emission mold; S2. Install an acoustic emission test platform, the acoustic emission test platform comprising: a mechanical loading device, an acoustic emission monitoring system and the acoustic emission mold produced in step S1; S3. Conduct multiple acoustic emission tests of the whole process of rock shear failure using the acoustic emission test platform installed in step S2, place the rock specimen processed in step S1 into the acoustic emission mold, arrange an acoustic emission probe in the acoustic emission mold, set the parameters of the acoustic emission monitoring system, and record the data of multiple acoustic emission tests; S4. Select three acoustic emission parameters with low correlation as calculation parameters of the rock shear failure whole process prediction model, and calculate the calculation parameters of the rock shear failure whole process prediction model based on the acoustic emission test data obtained in step S3; S5. Determine the input features and prediction target variables of the XGBoost algorithm, select data from the acoustic emission test data obtained in step S3, calculate the calculation parameters of the rock shear failure whole process prediction model selected in step S4, construct a data set for training the XGBoost algorithm, and train the XGBoost algorithm; S6. Initialize the XGBoost algorithm trained in step S5, determine the basic operation mode and operation parameters of the algorithm, and build a prediction model for the whole process of rock shear failure; S7. Input the calculation parameters of the rock shear failure whole process prediction method obtained in step S4 into the rock shear failure whole process prediction model constructed in step S6, and output the rock shear failure prediction result.

2. A method for predicting the entire process of rock shear failure based on acoustic emission monitoring according to claim 1, characterized in that: The rock specimen in step S1 is a cube specimen, the surface of the rock specimen is smooth, and the error of non-parallelism and non-verticality of the end is ±0.02mm; The acoustic emission mold also includes a wedge-shaped plug for fixing the acoustic emission probe.

3. A method for predicting the entire process of rock shear failure based on acoustic emission monitoring according to claim 2, characterized in that: The mechanical loading device in step S2 is arranged outside the acoustic emission mold to apply axial pressure and shear stress to the rock specimen; The acoustic emission monitoring system is arranged outside the acoustic emission mold, and includes an acoustic emission probe, a preamplifier, an acoustic emission acquisition card and acoustic emission processing software running on an electronic device, and is used to collect the acoustic emission transient waveform of the rock specimen in real time; The acoustic emission mold is arranged between the rock specimen and the mechanical loading device, and is used to carry the rock specimen and fix the acoustic emission probe; The installation spacing of the acoustic emission probes is 10-20 mm.

4. A method for predicting the entire process of rock shear failure based on acoustic emission monitoring according to claim 3, characterized in that: The step S3 comprises: Step 301. Place the rock specimen processed in step S1 into the acoustic emission mold, place the acoustic emission probe in contact with the rock specimen through the square hole of the acoustic emission mold, and insert the wedge-shaped plug into the square hole to fix the acoustic emission probe; Step 302. Set the acoustic emission monitoring system parameters, and perform mechanical preloading on the rock sample before the mechanical loading device performs formal mechanical loading. After the mechanical preloading is completed, the vertical force data, vertical displacement data, lateral force data, and lateral displacement data are reset to zero; Step 303. Load the normal force applied by the mechanical loading device to the initial normal stress value, and then use the mechanical loading device to perform mechanical loading at a shear rate of 0.2 mm / min. During the mechanical loading process, the mechanical loading process is kept synchronous with the data monitoring of the acoustic emission monitoring system. The initial normal stress value is set according to the particle size and normal stress conditions of the rock specimen; Step 304. When the mechanical loading in step 303 reaches a peak value, that is, the destruction phase of the acoustic emission test rock specimen ends, the mechanical loading of the mechanical loading device and the acoustic emission test data monitoring of the acoustic emission monitoring system are stopped at the same time, and the acoustic emission test data obtained by monitoring is collected and saved; Step 305: Take out the acoustic emission probe and the destroyed rock specimen, and clean the residue in the acoustic emission mold.

5. A method for predicting the entire process of rock shear failure based on acoustic emission monitoring according to claim 4, characterized in that: In step S301, the acoustic emission probes are evenly arranged at a position 5 cm above and below the rock specimen; The contact portion between the rock specimen and the acoustic emission probe is evenly coated with vaseline with a thickness of 0.1-0.2 mm.

6. A method for predicting the entire process of rock shear failure based on acoustic emission monitoring according to claim 5, characterized in that: The acoustic emission monitoring system parameters in step S302 include: number of channels, threshold value, amplification gain and sampling rate, the number of channels is set to 8, the threshold value is set to 45dB, the amplification gain is set to 40dB, and the sampling rate is set to 1MSPS; The preloading force of the mechanical preloading is 10%-20% of the expected maximum bearing capacity of the rock specimen.

7. A method for predicting the entire process of rock shear failure based on acoustic emission monitoring according to claim 6, characterized in that: The calculation parameters of the rock shear failure whole process prediction model in step S4 include: the logarithm of the acoustic emission cumulative energy, the b value and the main frequency A; The acoustic emission cumulative energy is as shown in formula (1): In the formula, E comulative It represents the cumulative energy of acoustic emission from time 0 to time T, E t Represents the energy carried by the acoustic emission signal at time t; The logarithm of the acoustic emission cumulative energy is Lg E comulative ; The b value is as shown in formula (2): In the formula, B represents the b value, m represents the total number of magnitude classes, and M j represents the median magnitude of the jth level, N j represents the number of acoustic emission events of the j-th magnitude, lg N as shown in formula (3): lg N=a-bM (3) In the formula, a and b are constants, M represents the magnitude as in formula (4), and N represents the total number of earthquakes with magnitude greater than M. M=(A dB +40) / 20 (4) In the formula, A dB represents the acoustic emission amplitude; The main frequency A is in KHz and is extracted from the acoustic emission signal in the acoustic emission test data collected in step S3 by using fast Fourier transform.

8. The method for predicting the entire process of rock shear failure based on acoustic emission monitoring according to claim 7 is characterized in that: The step S5 comprises: Step 501. Select data from the acoustic emission test data collected in step S3, calculate the logarithm of the acoustic emission cumulative energy, the b value and the main frequency A, and select the rock shear failure stage as the prediction target variable; Step 502: Divide the rock shear failure stage into a compaction stage, an elastic stage, an inelastic stage, and a post-peak failure stage, encode the compaction stage as 1, encode the elastic stage as 2, encode the inelastic stage as 3, and encode the post-peak failure as 4; Step 503. Perform corresponding rock shear failure stage encoding on the logarithm of the acoustic emission cumulative energy, b value and main frequency A of each rock specimen to form labeled data and construct a data set for training the XGBoost algorithm; Step 504: Divide the stress stage of the rock specimen into an elastic stage, a yield stage, a strengthening stage and a failure stage, and use the elastic stage, the yield stage, the strengthening stage and the failure stage as input features of the XGBoost algorithm; Step 505: Using the compaction stage, the elastic stage, the inelastic stage or the post-peak failure stage of the rock specimen as the target variable of the XGBoost algorithm; Step 506: Train the XGBoost algorithm.

9. A method for predicting the entire process of rock shear failure based on acoustic emission monitoring according to claim 8, characterized in that: The step S6 comprises: Step 601. Determine the basic operation mode, set the algorithm to run the data in the tree structure gbtree, build an algorithm similar to the tree structure to gradually analyze the data, and set general parameters; Step 602: Set the XGBoost algorithm boost parameters, where: The maximum number of iterations n_estimator is set to 100, The iteration step learning_rate is set to 0.7, The step size of each iteration is set to 0.

7. The maximum depth of the tree is set to 1.

5. The subsample ratio of random sampling for each tree is set to 1. Control the proportion of randomly sampled columns for each tree. Set colsample_bytree to 1. The proportion of column sampling colsample_bylevel is set to 1 for each node split in each tree. The weight L1 regularization term reg_alpha is set to 0, The weight L2 regularization term reg_lambda is set to 0, The minimum leaf node sample weight is set to 1. min_child_weight is set to 1; Step 603. Set the XGBoost algorithm learning target parameters, where: Objective is used to specify the logical relationship between the learning input features and the rock shear failure stage. The input features include the logarithm of the acoustic emission cumulative energy, b value and main frequency A. The XGBoost algorithm selects binary:logistic binary classification logistic regression, and the output is the probability that the rock specimen is in the unbroken, compaction stage, elastic stage, inelastic stage and post-peak failure stage. eval_metric is a quantitative evaluation of valid data. The XGBoost algorithm takes "auc", which is the area under the ROC curve, and is set to evaluate the performance of the model by calculating the area under the ROC curve of the receiver operating characteristic curve. seed is the random number seed, set to 1.

10. The method for predicting the entire process of rock shear failure based on acoustic emission monitoring according to claim 9, characterized in that: The general parameters set in step S601 include: turning on the silent mode silent to 0, outputting algorithm operation information; and setting the multi-thread control parameter nthread to the maximum number of threads on the computer.

Images