Rock shear failure whole process prediction method and system based on acoustic emission monitoring
By calibrating monitoring points on the rock surface, acquiring acoustic emission signals and calculating the contribution of shear components, dividing the data into grids and calculating the energy change rate, and combining this with a deep learning model, the problem of insufficient prediction accuracy during rock shear failure was solved, achieving accurate prediction of rock shear failure and location of high-risk areas.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- TAIYUAN UNIVERSITY OF TECHNOLOGY
- Filing Date
- 2026-01-21
- Publication Date
- 2026-06-19
AI Technical Summary
Existing technologies are insufficient for real-time prediction during rock shear failure, cannot locate microcracks or reflect internal rock changes, resulting in inaccurate predictions and failing to meet early warning requirements in complex engineering environments.
By calibrating monitoring points on the rock surface, acoustic emission signals are acquired, full waveform data is analyzed, shear component contribution is calculated, grids are divided and energy influence weights are calculated, a mapping relationship of cumulative energy change rate is established, and prediction is performed using a deep learning model.
It enables accurate prediction of rock shear failure, can locate high-risk areas, improves the timeliness and accuracy of prediction, and meets the real-time early warning needs in complex engineering environments.
Smart Images
Figure CN121540807B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of geotechnical engineering technology, specifically to a method and system for predicting the entire process of rock shear failure based on acoustic emission monitoring. Background Technology
[0002] In geotechnical engineering and mine safety, rock shear failure is a major cause of disasters such as rock instability, landslides, and tunnel collapses. Traditional monitoring methods rely heavily on macroscopic physical quantities such as displacement and stress, making it difficult to accurately capture and provide early warnings of the internal damage evolution process of rock. While acoustic emission monitoring has been introduced into existing technologies, most methods are limited to the statistical analysis of the number and energy of acoustic emission events, lacking in-depth analysis of the physical mechanisms of acoustic emission signals and failing to effectively combine spatiotemporal evolution characteristics for dynamic damage assessment. Furthermore, existing systems often struggle to achieve full automation and intelligence from signal acquisition to failure prediction, resulting in insufficient prediction accuracy and poor real-time performance, failing to meet the urgent need for early warning of the entire rock shear failure process in complex engineering environments.
[0003] In the prior art, CN119534637A discloses a method for predicting the entire process of rock shear failure based on acoustic emission monitoring. This method involves collecting rock samples and processing them into multiple rock specimens, fabricating an acoustic emission mold, installing an acoustic emission test platform, conducting multiple acoustic emission tests on the entire process of rock shear failure, recording the data from these tests, selecting and calculating the computational parameters of the prediction model based on the experimental data, and finally constructing the prediction model based on the XGBoost algorithm. The computational parameters of the prediction method are then input into the model, and the predicted rock shear failure results are output.
[0004] The main problems with the above scheme are: using the final, completed data from multiple experiments as the training set, it is impossible to predict the future state of a single rock sample in real time based on the evolution trend of the current data stream during the failure process; it is difficult to locate the precise location and fracture type of microcracks inside the rock, resulting in the loss of spatial dimension information of damage evolution; and it cannot reflect the changes inside the rock during the entire shear failure process, leading to reduced prediction accuracy.
[0005] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention
[0006] The purpose of this invention is to provide a method and system for predicting the entire process of rock shear failure based on acoustic emission monitoring, so as to solve the problems mentioned in the background art.
[0007] To achieve the above objectives, the present invention provides the following technical solution:
[0008] A method for predicting the entire process of rock shear failure based on acoustic emission monitoring, comprising the following steps:
[0009] Step 1: Mark several monitoring points on the surface of the rock to be tested and determine their three-dimensional coordinates. Acoustic emission signals are acquired at each monitoring point simultaneously. The acoustic emission signals of all monitoring points are analyzed, and the full waveform data of each acoustic emission event is determined by combining the acoustic emission signals of all monitoring points.
[0010] Step 2: Based on the full waveform data of the acoustic emission event, obtain the arrival time of the P wave at different monitoring points, and combine it with the three-dimensional coordinates of the corresponding monitoring points to determine the three-dimensional coordinates of the occurrence point of each acoustic emission event; perform moment tensor inversion based on the full waveform data of each acoustic emission event, and decompose the moment tensor to calculate the shear component contribution of each acoustic emission event.
[0011] Step 3: Divide the rock to be tested into several grids. For each acoustic emission event point, calculate its energy influence weight on each grid, and then calculate its shear energy contribution to each grid.
[0012] Step 4: Divide the monitoring time of acoustic emission events into several time windows of equal length. For each grid, generate the cumulative energy of each time window based on the shear energy contribution of different acoustic emission events, and then generate the rate of change of the cumulative energy of the grid in each time window. Establish the mapping relationship between the rate of change of the cumulative energy of three consecutive time windows and the rate of change of the cumulative energy of the next adjacent time window.
[0013] Step 5: Determine the rate of change of cumulative energy when the rock undergoes shear failure, set it as the failure rate threshold, obtain the rate of change of cumulative energy in the next time window based on the rate of change of cumulative energy in three consecutive time windows, and determine whether shear failure will occur in the next time window by combining the rate of change of cumulative energy in the next time window with the failure rate threshold.
[0014] Furthermore, the principle underlying the determination of the full waveform data for each acoustic emission event is as follows:
[0015] For each acoustic emission event, the monitoring point that first detects the acoustic emission signal of that event is selected as the main trigger monitoring point, and a trigger time window is set. From the moment the acoustic emission signal of the acoustic emission event is received by the main trigger monitoring point until the end of a trigger time window, the acoustic emission signals detected by all other monitoring points during this period are searched. All such acoustic emission signals together constitute the complete acoustic emission signal of an acoustic emission event. Based on the complete acoustic emission signal of each acoustic emission event, its full waveform data is determined. The full waveform data includes signal energy, signal duration, and P-wave waveform.
[0016] Furthermore, the principle behind setting the trigger time window is as follows:
[0017] By iterating through all monitoring points and calculating the Euclidean distance between each pair of monitoring points, the maximum time required for the acoustic emission signal to propagate between the monitoring points can be calculated. The formula used is as follows:
[0018]
[0019] in, This indicates the longest time required for an acoustic emission signal to propagate between monitoring points. Indicates the maximum distance between monitoring points. express The speed at which waves propagate in rock;
[0020] The longest time required for the acoustic emission signal to propagate between monitoring points is used as the trigger time window length.
[0021] Furthermore, the principle underlying the calculation of the shear component contribution for each acoustic emission event is as follows:
[0022] The three-dimensional coordinates of each acoustic emission event point are calculated based on the Geiger algorithm.
[0023] The moment tensor is:
[0024]
[0025] in, Represents the moment tensor. This represents the moment of a couple acting in a plane perpendicular to the x-axis. Indicates action on the vertical Moment of couple in the axial plane Indicates action on the vertical Moment of couple in the axial plane express Pure shear couple moment in the plane express Pure shear couple moment in the plane express Pure shear couple moment in the plane;
[0026] Decomposing the moment tensor into its isotropic and partial tensor components, the moment tensor is expressed as:
[0027]
[0028] in, Indicates the isotropic part, Indicates the partial tensor component;
[0029] The partial tensor part is:
[0030]
[0031]
[0032]
[0033] in, Represents the trace of the moment tensor. Represents the identity matrix;
[0034] Based on eigenvalue decomposition, the partial tensor part is decomposed into a double-couple part and a compensated linear vector dipole part:
[0035]
[0036] in, Indicates the double couple part. Represents the compensating linear vector dipole component;
[0037] The formula for calculating the contribution of the shear component is:
[0038]
[0039] in, Indicates the contribution of shear components. This represents the Frobenius norm.
[0040] Furthermore, the principle of dividing the rock to be tested into several grids is as follows:
[0041] Based on the deployment range of monitoring points, the three-dimensional spatial boundary of the rock to be detected is determined, and then the boundaries of the rock in the x, y, and z directions are determined. The region within the boundary is then uniformly divided into several areas of size [missing information]. Grid, in which, Indicates the grid size. This represents the length of the grid in the x-direction. This represents the length of the grid in the y-direction. This represents the length of the grid in the z-direction; the grid size is determined based on the positioning accuracy and monitoring requirements of acoustic emission events, using an expert scoring method.
[0042] Furthermore, the principle underlying the calculation of the shear energy contribution of each acoustic emission event point to each grid is as follows:
[0043] The formula for calculating the weight of energy influence is:
[0044]
[0045]
[0046] in, Indicates the first From the point where the acoustic emission event occurred to the... The Euclidean distance between the center points of each grid An index representing an acoustic emission event. Indicates the grid index, They represent the first The horizontal, vertical, and axial coordinates of the location where the acoustic emission event occurred. Indicates the first The horizontal, vertical, and axial coordinates of the center point of each grid. Indicates the first The point where the acoustic emission event occurs is related to the first The energy at each grid center point affects the weight. Indicates the attenuation coefficient;
[0047] The formula for calculating the shear energy contribution is:
[0048]
[0049] in, Indicates the first The acoustic emission event affects the first Shear energy contribution of each grid, Indicates the first The energy produced by an acoustic emission event Indicates the first The contribution of shearing components to each acoustic emission event.
[0050] Furthermore, the principle underlying the rate of cumulative energy change of the generated mesh in each time window is as follows:
[0051] The formula used to calculate the cumulative energy of the grid over each time window is:
[0052]
[0053] in, Indicates the first Within the first time window The cumulative energy of each grid, Indicates the first The number of acoustic emission events within a time window. Indicates the index of the time window. Indicates the first Within the first time window The acoustic emission event affects the first Shear energy contribution of each grid;
[0054] The formula used to calculate the rate of change of cumulative energy is:
[0055]
[0056] in, Indicates the first Within the first time window The rate of cumulative energy change in each grid, Indicates the first The first acoustic emission event within the time window affects the first... The shear energy contribution of the first grid, i.e., the first grid at the time of the first acoustic emission event. The cumulative energy of each grid, Indicates the first The first acoustic emission event occurred within a certain time window, and the first... The time interval between infrasound emission events.
[0057] Furthermore, the principle for establishing the mapping relationship between the cumulative energy change rate of three consecutive time windows and the cumulative energy change rate of the next adjacent time window is as follows:
[0058] A deep learning model is constructed by dividing the historical monitoring data into several time windows according to the monitoring time. Each group of four consecutive time windows is used as a set, and the cumulative energy change rate of each time window is calculated. The cumulative energy change rates of the first three time windows are used as inputs, and the cumulative energy change rate of the fourth time window is used as a label to train the change prediction model.
[0059] Furthermore, the principle underlying the determination of whether shear failure will occur in the next time window is as follows:
[0060] The cumulative energy change rate of the mesh, collected and calculated over three consecutive time windows, is input into the change prediction model to obtain the cumulative energy change rate for the next time window. If the cumulative energy change rate of the next time window is greater than or equal to the failure rate threshold, the mesh will undergo shear failure in the next time window; otherwise, shear failure will not occur.
[0061] This invention also provides a rock shear failure prediction system based on acoustic emission monitoring. The system is used to implement the aforementioned rock shear failure prediction method based on acoustic emission monitoring, specifically including:
[0062] The data acquisition and optimization module is used to mark several monitoring points on the surface of the rock to be tested and determine their three-dimensional coordinates. It simultaneously acquires acoustic emission signals at each monitoring point, analyzes the acoustic emission signals of all monitoring points, and combines the acoustic emission signals of all monitoring points to determine the full waveform data of each acoustic emission event.
[0063] The shearing calculation module is used to obtain the arrival time of the P wave at different monitoring points based on the full waveform data of the acoustic emission event, and determine the three-dimensional coordinates of the occurrence point of each acoustic emission event by combining the three-dimensional coordinates of the corresponding monitoring points; it performs moment tensor inversion based on the full waveform data of each acoustic emission event, and decomposes the moment tensor to calculate the shearing component contribution of each acoustic emission event.
[0064] The energy contribution calculation module is used to divide the rock to be detected into several grids, calculate the energy influence weight of each acoustic emission event point on each grid, and then calculate its shear energy contribution to each grid.
[0065] The energy change calculation module is used to divide the monitoring time of acoustic emission events into several time windows of equal length. For each grid, the cumulative energy of each time window is generated based on the shear energy contribution of different acoustic emission events, and then the cumulative energy change rate of the grid in each time window is generated. The mapping relationship between the cumulative energy change rate of three consecutive time windows and the cumulative energy change rate of the next adjacent time window is established.
[0066] The comprehensive judgment module is used to determine the rate of cumulative energy change when rock undergoes shear failure, which is set as the failure rate threshold. Based on the rate of cumulative energy change of three consecutive time windows, the rate of cumulative energy change of the next time window is obtained, and combined with the failure rate threshold, it is determined whether shear failure occurs in the next time window.
[0067] Compared with the prior art, the beneficial effects of the present invention are:
[0068] This invention ensures that a complete waveform signal of an acoustic emission event can be spliced from the signals received from all monitoring points by setting a main trigger monitoring point and a trigger time window, thus ensuring the complete acquisition of full waveform data for each acoustic emission event. Through moment tensor inversion, the shear component contribution of each acoustic emission event is quantitatively calculated, clearly determining the extent to which a micro-fracture event is shear failure, thereby accurately screening out micro-fracture signals directly related to the final shear instability of the rock and effectively filtering out other types of interference signals.
[0069] This invention further utilizes a three-dimensional meshing of the rock under test, combined with a spatial attenuation weight model, to distribute the shear energy of each acoustic emission event to each mesh according to spatial weights. This processing method not only reflects the real physical process of energy attenuation with distance but also realizes the continuous spatial distribution of damage accumulation. It can intuitively display the local concentrated areas of damage within the rock, providing a spatial dimension basis for predicting subsequent failure locations. The invention calculates the rate of change of accumulated energy in each mesh according to time windows, dynamically describing the accumulation process and rate of change of damage energy within each mesh. By anticipating these temporal characteristics, the prediction is not only based on the total accumulated amount but also focuses on the dynamic trend of damage evolution, enhancing the sensitivity of the scheme to precursors of failure and the timeliness of early warning. The predicted rate of change of accumulated energy is compared with a preset failure rate threshold to determine whether shear failure will occur in each mesh in the next time window. This not only achieves the prediction of failure time but also accurately locates high-risk areas. Attached Figure Description
[0070] Figure 1 This is a schematic diagram of the method flow of an embodiment of the present invention;
[0071] Figure 2 This is a schematic diagram of the fitting curve of the cumulative energy changing over time in an embodiment of the present invention;
[0072] Figure 3 This is a schematic diagram of the system modules in an embodiment of the present invention. Detailed Implementation
[0073] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.
[0074] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.
[0075] Example:
[0076] Please see Figures 1 to 3The present invention provides a technical solution:
[0077] A method for predicting the entire process of rock shear failure based on acoustic emission monitoring, comprising the following steps:
[0078] Step 1: Mark several monitoring points on the surface of the rock to be tested and determine their three-dimensional coordinates. Acoustic emission signals are acquired at each monitoring point simultaneously. The acoustic emission signals of all monitoring points are analyzed, and the full waveform data of each acoustic emission event is determined by combining the acoustic emission signals of all monitoring points.
[0079] In this embodiment, the principle upon which the full waveform data of each acoustic emission event is determined is as follows:
[0080] For each acoustic emission event, the monitoring point that first detects the acoustic emission signal of that event is selected as the main trigger monitoring point, and a trigger time window is set. From the moment the acoustic emission signal of the acoustic emission event is received by the main trigger monitoring point until the end of a trigger time window, the acoustic emission signals detected by all other monitoring points during this period are searched. All such acoustic emission signals together constitute the complete acoustic emission signal of an acoustic emission event. Based on the complete acoustic emission signal of each acoustic emission event, its full waveform data is determined. The full waveform data includes signal energy, signal duration, and P-wave waveform.
[0081] Monitoring points are evenly distributed across the rock surface, covering the rock to be tested, ensuring effective capture of acoustic emission signals from different locations within the rock. Acoustic emission signals from within the rock represent transient elastic waves released during deformation and fracturing. The three-dimensional coordinates of each monitoring point are obtained using a total station, and an acoustic emission sensor is installed at each point. When an acoustic emission event occurs, the signal propagates outwards from the point of origin in the form of P-waves, S-waves, etc. Due to the different sensor locations, the arrival time of the same signal at each monitoring point varies. Setting a trigger time window ensures waveform integrity; when a monitoring point first detects an acoustic emission signal, subsequent... Within a brief time window, other monitoring points will also detect this acoustic emission signal. The signals detected by all monitoring points constitute the complete acoustic emission signal. The integrity of each acoustic emission signal is ensured by triggering a time window. The complete acoustic emission signal is a multi-channel time series, with each channel corresponding to the signal collected by a monitoring point. The signal detected by the main triggering monitoring point is used as the time reference. The full waveform data mainly includes: signal energy, which reflects the amount of energy released by the acoustic emission event; signal duration, which represents the time from the initial arrival of the P-wave to the signal decaying to the background noise level; and the P-wave waveform, which is the original waveform data of the P-band, used for positioning in the moment tensor inversion.
[0082] The principle behind setting a trigger time window is as follows:
[0083] By iterating through all monitoring points and calculating the Euclidean distance between each pair of monitoring points, the maximum time required for the acoustic emission signal to propagate between the monitoring points can be calculated. The formula used is as follows:
[0084]
[0085] in, This indicates the longest time required for an acoustic emission signal to propagate between monitoring points. Indicates the maximum distance between monitoring points. express The speed at which waves propagate in rock;
[0086] The longest time required for the acoustic emission signal to propagate between monitoring points is used as the trigger time window length.
[0087] The trigger time window is a time interval used to determine whether signals from the same acoustic emission event at different monitoring points belong to the same event. The acoustic emission signal propagates from the event origin to each monitoring point in the form of an elastic wave. Since the distances between each monitoring point and the event origin are different, the arrival time of the signal at each monitoring point also varies. To ensure that all monitoring points receive the signal of the same event, a time window needs to be set. Its length should not be less than the time required for the signal to propagate from the event origin to the farthest monitoring point. This time window is the trigger time window. If the window is too short, signals from some monitoring points may be missed, preventing the formation of complete waveform data; if the window is too long, signals from different events may be incorrectly classified as the same event. Using this as the window length ensures that all relevant signals are included, while avoiding event confusion caused by excessively long windows.
[0088] Step 2: Based on the full waveform data of the acoustic emission event, obtain the arrival time of the P wave at different monitoring points, and combine it with the three-dimensional coordinates of the corresponding monitoring points to determine the three-dimensional coordinates of the occurrence point of each acoustic emission event; perform moment tensor inversion based on the full waveform data of each acoustic emission event, and decompose the moment tensor to calculate the shear component contribution of each acoustic emission event.
[0089] In this embodiment, the principle underlying the calculation of the shear component contribution for each acoustic emission event is as follows:
[0090] The three-dimensional coordinates of each acoustic emission event point are calculated based on the Geiger algorithm.
[0091] The Geiger algorithm is a seismic event location method based on iterative least squares, which infers the location of the event by using the arrival time of P waves received by multiple sensors. The technical features of location using the Geiger algorithm have been disclosed in the article "Rock Acoustic Emission Source Location Method Based on Full-Phase Geiger Algorithm, November 2016, Vol. 29, No. 11, Journal of Sensor Technology, Huang Xiaohong, Sun Guoqing, Zhang Kaiyue", and will not be repeated here.
[0092] The moment tensor is:
[0093]
[0094] in, Represents the moment tensor. This represents the moment of a couple acting in a plane perpendicular to the x-axis. Indicates action on the vertical Moment of couple in the axial plane Indicates action on the vertical Moment of couple in the axial plane express Pure shear couple moment in the plane express Pure shear couple moment in the plane express Pure shear couple moment in the plane;
[0095] The moment tensor is a 3×3 symmetric matrix used to describe the equivalent force system at the point of occurrence of an acoustic emission event. Acoustic emission events produce microcracks, and the rock masses on both sides will undergo relative displacement. It can be equivalent to a set of two couple force systems occurring at the point of occurrence. The value of the moment tensor reflects the magnitude and mechanism type of the acoustic emission event. The mechanism types include pure shear, pure tension, and mixed type. This represents a tension or compression dipole along the x-axis. It describes the direction of the force couple along the x-axis and the surface of action perpendicular to the x-axis. A positive value indicates tension along the x-axis, and a negative value indicates compression along the x-axis. This indicates a tensile or compressive dipole along the y-axis, describing the direction of the couple along the y-axis and the surface of action perpendicular to the y-axis. This indicates a tension or compression dipole along the z-axis, describing the direction of the couple along the z-axis and the surface of action perpendicular to the z-axis. express The pure shear couple moment in a plane describes a pair of forces whose direction is along the x-axis but whose plane of action is perpendicular to the y-axis. It is directly related to shear fracture and describes... Similarly, shearing in a plane... , They described respectively , Shearing action in a plane;
[0096] Decomposing the moment tensor into its isotropic and partial tensor components, the moment tensor is expressed as:
[0097]
[0098] in, Indicates the isotropic part, Indicates the partial tensor component;
[0099] The isotropic part represents fractures related to volume changes, corresponding to tensile or compressive fractures, while the partial tensor part represents fractures related to shape changes, corresponding to shear fractures. Any second-order symmetric tensor can be uniquely decomposed into an isotropic tensor. and a partial tensor , This reflects the volume change caused by the rupture. It reflects the shape changes caused by fracture. This decomposition can distinguish fracture types and quantify the contribution of shear components.
[0100] The partial tensor part is:
[0101]
[0102]
[0103]
[0104] in, Represents the trace of the moment tensor. Represents the identity matrix;
[0105] The partial tensor describes the anisotropic component of the moment tensor and is calculated based on the trace decomposition of the tensor. Its purpose is to remove volume change components from the original moment tensor, retaining only pure shape change information. The trace reflects the rate of volume change; the partial tensor only describes shape changes and does not cause volume changes, therefore its trace is zero. The trace of the partial tensor only considers the isotropic component. The isotropic component is a diagonal matrix with identical values in all directions, and the values on its diagonal are equal to the sum of the three main diagonal elements of the moment tensor. Off-diagonal elements are zero; let ,but , Therefore , By removing the isotropic portion from the moment tensor, we obtain the partial tensor portion.
[0106] Based on eigenvalue decomposition, the partial tensor part is decomposed into a double-couple part and a compensated linear vector dipole part:
[0107]
[0108] in, Indicates the double couple part. Represents the compensating linear vector dipole component;
[0109] The deviator tensor has eliminated isotropic variations and only describes shape changes; the deviator tensor can be further decomposed into a double couple part and a compensated linear vector dipole part. The double couple part corresponds to pure shear fracture, and the compensated linear vector dipole part represents a hybrid mechanism between shear and tension. Given a symmetric matrix, eigenvalue decomposition yields three eigenvalues. ,and The moment tensor decomposition yields The eigenvalues of the two-couple part satisfy Corresponding to the pure shear fracture mechanism, the eigenvalues of the compensating linear vector dipole portion satisfy... , :, corresponding to a hybrid fracture mechanism; ;in, express The corresponding feature vector, express The corresponding feature vector.
[0110] The formula for calculating the contribution of the shear component is:
[0111]
[0112] in, Indicates the contribution of shear components. This represents the Frobenius norm.
[0113] Indicates pure shearing component, This represents the total magnitude of the entire moment tensor. The larger the value, the closer the acoustic emission event is to pure shear fracture, and the greater the contribution of the shear component. The contribution of the shear component is proportional to the Frobenius norm of the pure shear component.
[0114] Step 3: Divide the rock to be tested into several grids. For each acoustic emission event point, calculate its energy influence weight on each grid, and then calculate its shear energy contribution to each grid.
[0115] In this embodiment, the principle of dividing the rock to be tested into several grids is as follows:
[0116] Based on the deployment range of monitoring points, the three-dimensional spatial boundary of the rock to be detected is determined, and then the boundaries of the rock in the x, y, and z directions are determined. The region within the boundary is then uniformly divided into several areas of size [missing information]. Grid, in which, Indicates the grid size. This represents the length of the grid in the x-direction. This represents the length of the grid in the y-direction. This represents the length of the grid in the z-direction; the grid size is determined based on the positioning accuracy and monitoring requirements of acoustic emission events, using an expert scoring method.
[0117] When determining the grid size, the accuracy of acoustic emission event location, monitoring resolution, and computational efficiency are comprehensively considered. Each candidate size scheme is scored by an expert scoring method, and the scheme with the highest comprehensive score is selected as the final grid size.
[0118] The principle underlying the calculation of the shear energy contribution of each acoustic emission event point to each grid is as follows:
[0119] The formula for calculating the weight of energy influence is:
[0120]
[0121]
[0122] in, Indicates the first From the point where the acoustic emission event occurred to the... The Euclidean distance between the center points of each grid An index representing an acoustic emission event. Indicates the grid index, They represent the first The horizontal, vertical, and axial coordinates of the location where the acoustic emission event occurred. Indicates the first The horizontal, vertical, and axial coordinates of the center point of each grid. Indicates the first The point where the acoustic emission event occurs is related to the first The energy at each grid center point affects the weight. Indicates the attenuation coefficient;
[0123] The energy influence weight reflects the degree of energy contribution of acoustic emission events to grids at different locations in space. The energy released by acoustic emission events propagates in the rock medium in the form of elastic waves, and its energy attenuates with increasing distance. This attenuation coefficient is used to determine the energy contribution. The attenuation rate is controlled and adjusted according to factors such as rock type and medium homogeneity to adapt to different geological conditions and monitoring scenarios. The weight value decreases with increasing distance, reflecting the attenuation characteristics of acoustic emission energy during propagation. Grids closer to the event point receive greater energy influence. By assigning a weight to each grid, the energy released by a single acoustic emission event can be distributed spatially, thus more realistically simulating the energy accumulation process within the rock. The weighting mechanism ensures continuity in energy accumulation between adjacent grids, reflecting the spatial continuity of rock damage evolution. By distributing the energy of each event to each grid according to its weight, the spatial distribution of the energy field can be constructed, allowing for the calculation of the accumulated energy of each grid and its change over time. ,like This indicates that the weight decays slowly, and acoustic emission events still have a significant impact on distant meshes, making it suitable for uniform, low-attenuation media; if The weight decays sharply, and the energy of the acoustic emission event is highly concentrated near the point of occurrence of the acoustic emission event. It is suitable for highly attenuated or non-uniform media.
[0124] The formula for calculating the shear energy contribution is:
[0125]
[0126] in, Indicates the first The acoustic emission event affects the first Shear energy contribution of each grid, Indicates the first The energy produced by an acoustic emission event Indicates the first The contribution of shearing components to each acoustic emission event.
[0127] Shear energy contribution reflects the effective shear energy of a single acoustic emission event in a specific grid. It considers not only the total energy of the event but also the influence of its shear mechanism components and spatial location. It is a spatially weighted pure shear energy allocation value. Grids with high shear energy indicate that shear-type microfractures are active in that region and are potential damage concentration areas. It reflects the non-uniform distribution and localization trend of damage in space. As multiple acoustic emission events are superimposed, the cumulative shear energy of each grid will gradually increase, reflecting the gradual accumulation process of damage. This represents the total energy released by the acoustic emission event, reflecting the size of the microfracture. Higher energy indicates stronger elastic wave energy released by the acoustic emission event, and a greater contribution to the damage to the surrounding medium; shear component contribution... The larger the value, the closer the acoustic emission event is to pure shear fracture; the shear energy contribution is proportional to the energy of the acoustic emission event, proportional to the shear component contribution of the acoustic emission event, and proportional to the weight of the energy influence of the acoustic emission event on the mesh.
[0128] Step 4: Divide the monitoring time of acoustic emission events into several time windows of equal length. For each grid, generate the cumulative energy of each time window based on the shear energy contribution of different acoustic emission events, and then generate the rate of change of the cumulative energy of the grid in each time window. Establish the mapping relationship between the rate of change of the cumulative energy of three consecutive time windows and the rate of change of the cumulative energy of the next adjacent time window.
[0129] In this embodiment, the principle underlying the generation of the grid's cumulative energy change rate across each time window is as follows:
[0130] The formula used to calculate the cumulative energy of the grid over each time window is:
[0131]
[0132] in, Indicates the first Within the first time window The cumulative energy of each grid, Indicates the first The number of acoustic emission events within a time window. Indicates the index of the time window. Indicates the first Within the first time window The acoustic emission event affects the first Shear energy contribution of each grid;
[0133] Accumulated energy Reflects the first The cumulative shear loss of rock within a specific grid area within a time window. The higher the cumulative energy, the more or stronger the shear-type micro-fracture events experienced in that area within that time window, the more significant the damage accumulation, and the more likely these areas are to undergo shear failure. The cumulative energy reflects the entire damage development process within a time window; it is the sum of the shear energy of all acoustic emission events acting on the grid within that time window, and is directly proportional to the number of acoustic emission events acting on the grid and the shear energy.
[0134] Table 1 reflects the change of accumulated energy of a grid over time. In the initial stage, microcracks are randomly distributed. Under low stress, internal microcracks randomly initiate and propagate. Acoustic emission events are scattered and have low energy, resulting in a slow accumulation growth. In the middle stage, microcracks gradually increase and begin to interact, and the energy accumulation rate gradually accelerates, but remains relatively stable. In the later stage, microcracks gradually converge and connect, forming macroscopic fracture zones. Local stress concentration is significantly enhanced, the frequency of acoustic emission events increases, and the accumulated energy increases sharply.
[0135] Table 1. Cumulative Energy Variation over Time
[0136]
[0137] The formula used to calculate the rate of change of cumulative energy is:
[0138]
[0139] in, Indicates the first Within the first time window The rate of cumulative energy change in each grid, Indicates the first The first acoustic emission event within the time window affects the first... The shear energy contribution of the first grid, i.e., the first grid at the time of the first acoustic emission event. The cumulative energy of each grid, Indicates the first The first acoustic emission event occurred within a certain time window, and the first... The time interval between infrasound emission events.
[0140] The rate of change of accumulated energy reflects the rate of damage accumulation. The greater the rate of change, the more shear loss energy is added to the grid area per unit time, which means that the micro-fracture activity in the area is accelerating. If the rate of change of accumulated energy in different time windows continues to increase, it means that the loss in the area is accelerating, which may indicate instability or destruction. If the rate of change of accumulated energy in different time windows tends to stabilize or decrease, it means that the loss is developing slowly or tending to stop. This reflects the total damage energy accumulated in the mesh due to shear energy from acoustic emission events within the time window. The first acoustic emission time contributes to the shear energy of the mesh, and the second time represents the contribution of the first acoustic emission time to the shear energy of the mesh. The energy in the initial state for each time window is such that the rate of change in each time window reflects the pure change from the first acoustic emission event to the current moment. It represents the net increase in cumulative shear energy per unit time;
[0141] The principle for establishing the mapping relationship between the cumulative energy change rate of three consecutive time windows and the cumulative energy change rate of the next adjacent time window is as follows:
[0142] A deep learning model is constructed by dividing the historical monitoring data into several time windows according to the monitoring time. Each group of four consecutive time windows is used as a set, and the cumulative energy change rate of each time window is calculated. The cumulative energy change rate of the first three time windows is used as the label, and the cumulative energy change rate of the fourth time window is used as the label to train the change prediction model.
[0143] From historical monitoring data, several time windows are divided according to time. Every four consecutive time windows are grouped together. The cumulative energy change rate of the grid within each time window is calculated. The cumulative energy change rates of the first three time windows in each group are used as reference values, and the cumulative energy change rate of the fourth time window is used as the target value. The correspondence between several sets of reference values and target values is obtained. These data are divided into training set and validation set in an 8:2 ratio. The reference value is used as input and the target value is used as label to train the change prediction model. The validation set is used for validation. The mean squared error function is used as the loss function. The initial learning rate is set to 0.0001. After every 30 training iterations, the learning rate is multiplied by 0.95 to ensure stable convergence in the later stages of training. Training is stopped when the loss function of the validation set no longer decreases. The model at this point is saved as the change prediction model.
[0144] Step 5: Determine the rate of change of cumulative energy when the rock undergoes shear failure, set it as the failure rate threshold, obtain the rate of change of cumulative energy in the next time window based on the rate of change of cumulative energy in three consecutive time windows, and determine whether shear failure will occur in the next time window by combining the rate of change of cumulative energy in the next time window with the failure rate threshold.
[0145] In this embodiment, the principle used to determine whether shearing failure occurs in the next time window is as follows:
[0146] The cumulative energy change rate of the mesh, collected and calculated over three consecutive time windows, is input into the change prediction model to obtain the cumulative energy change rate for the next time window. If the cumulative energy change rate of the next time window is greater than or equal to the failure rate threshold, the mesh will undergo shear failure in the next time window; otherwise, shear failure will not occur.
[0147] A higher rate of cumulative energy change indicates a higher risk of grid failure. When the rate of cumulative energy change reaches a certain threshold, it indicates that the rock in the grid area will undergo shear failure. This threshold, determined experimentally, is set as the failure rate threshold. Based on the known cumulative energy change rates of three consecutive time windows, the cumulative energy change rate of the unknown next time window is predicted. The cumulative energy change rate of the next time window is compared with the failure rate threshold. If the cumulative energy change rate of the next time window is greater than or equal to the failure rate threshold, the grid is considered to undergo shear failure in the next time window; otherwise, shear failure will not occur. Prediction on a grid-by-grid basis allows for the location of specific areas, identifying high-risk areas before actual failure occurs.
[0148] Please see Figure 3 The present invention also provides a rock shear failure prediction system based on acoustic emission monitoring. The system is used to implement the above-mentioned rock shear failure prediction method based on acoustic emission monitoring, specifically including:
[0149] The data acquisition and optimization module is used to mark several monitoring points on the surface of the rock to be tested and determine their three-dimensional coordinates. It simultaneously acquires acoustic emission signals at each monitoring point, analyzes the acoustic emission signals of all monitoring points, and combines the acoustic emission signals of all monitoring points to determine the full waveform data of each acoustic emission event.
[0150] The shearing calculation module is used to obtain the arrival time of the P wave at different monitoring points based on the full waveform data of the acoustic emission event, and determine the three-dimensional coordinates of the occurrence point of each acoustic emission event by combining the three-dimensional coordinates of the corresponding monitoring points; it performs moment tensor inversion based on the full waveform data of each acoustic emission event, and decomposes the moment tensor to calculate the shearing component contribution of each acoustic emission event.
[0151] The energy contribution calculation module is used to divide the rock to be detected into several grids, calculate the energy influence weight of each acoustic emission event point on each grid, and then calculate its shear energy contribution to each grid.
[0152] The energy change calculation module is used to divide the monitoring time of acoustic emission events into several time windows of equal length. For each grid, the cumulative energy of each time window is generated based on the shear energy contribution of different acoustic emission events, and then the cumulative energy change rate of the grid in each time window is generated. The mapping relationship between the cumulative energy change rate of three consecutive time windows and the cumulative energy change rate of the next adjacent time window is established.
[0153] The comprehensive judgment module is used to determine the rate of cumulative energy change when rock undergoes shear failure, which is set as the failure rate threshold. Based on the rate of cumulative energy change of three consecutive time windows, the rate of cumulative energy change of the next time window is obtained, and combined with the failure rate threshold, it is determined whether shear failure occurs in the next time window.
[0154] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.
[0155] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.
[0156] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.
[0157] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.
Claims
1. A method for predicting the entire process of rock shear failure based on acoustic emission monitoring, characterized in that, The specific steps include: Step 1: Mark several monitoring points on the surface of the rock to be tested and determine their three-dimensional coordinates. Acoustic emission signals are acquired at each monitoring point simultaneously. The acoustic emission signals of all monitoring points are analyzed, and the full waveform data of each acoustic emission event is determined by combining the acoustic emission signals of all monitoring points. Step 2: Based on the full waveform data of the acoustic emission event, obtain the arrival time of the P wave at different monitoring points, and combine it with the three-dimensional coordinates of the corresponding monitoring points to determine the three-dimensional coordinates of the occurrence point of each acoustic emission event; perform moment tensor inversion based on the full waveform data of each acoustic emission event, and decompose the moment tensor to calculate the shear component contribution of each acoustic emission event. Step 3: Divide the rock to be tested into several grids. For each acoustic emission event point, calculate its energy influence weight on each grid, and then calculate its shear energy contribution to each grid. Step 4: Divide the monitoring time of acoustic emission events into several time windows of equal length. For each grid, generate the cumulative energy of each time window based on the shear energy contribution of different acoustic emission events, and then generate the rate of change of the cumulative energy of the grid in each time window. Establish the mapping relationship between the rate of change of the cumulative energy of three consecutive time windows and the rate of change of the cumulative energy of the next adjacent time window. Step 5: Determine the rate of change of cumulative energy when the rock undergoes shear failure, set it as the failure rate threshold, obtain the rate of change of cumulative energy in the next time window based on the rate of change of cumulative energy in three consecutive time windows, and combine it with the failure rate threshold to determine whether shear failure occurs in the next time window. The principle underlying the calculation of the shear energy contribution of each acoustic emission event point to each grid is as follows: The formula for calculating the weight of energy influence is: in, Indicates the first From the point where the acoustic emission event occurred to the... Euclidean distance between the center points of each grid An index representing an acoustic emission event. Indicates the grid index, They represent the first The horizontal, vertical, and axial coordinates of the location where the acoustic emission event occurred. Indicates the first The horizontal, vertical, and axial coordinates of the center point of each grid. Indicates the first The point where the acoustic emission event occurred was related to the first... The energy at the center point of each grid affects the weight. Indicates the attenuation coefficient; The formula for calculating the shear energy contribution is: in, Indicates the first The acoustic emission event affects the first Shear energy contribution of each grid, Indicates the first The energy produced by an acoustic emission event Indicates the first The contribution of shearing components to each acoustic emission event.
2. The method for predicting the entire process of rock shear failure based on acoustic emission monitoring according to claim 1, characterized in that: The principle upon which the full waveform data for each acoustic emission event is determined in step 1 is as follows: For each acoustic emission event, the monitoring point that first detects the acoustic emission signal of that event is selected as the main trigger monitoring point, and a trigger time window is set. From the moment the acoustic emission signal of the acoustic emission event is received by the main trigger monitoring point until the end of a trigger time window, the acoustic emission signals detected by all other monitoring points during this period are searched. All such acoustic emission signals together constitute the complete acoustic emission signal of an acoustic emission event. Based on the complete acoustic emission signal of each acoustic emission event, its full waveform data is determined. The full waveform data includes signal energy, signal duration, and P-wave waveform.
3. The method for predicting the entire process of rock shear failure based on acoustic emission monitoring according to claim 2, characterized in that: The principle behind setting a trigger time window is as follows: By iterating through all monitoring points and calculating the Euclidean distance between each pair of monitoring points, the maximum time required for the acoustic emission signal to propagate between the monitoring points can be calculated. The formula used is as follows: in, This indicates the longest time required for an acoustic emission signal to propagate between monitoring points. Indicates the maximum distance between monitoring points. express The speed at which waves propagate in rock; The longest time required for the acoustic emission signal to propagate between monitoring points is used as the trigger time window length.
4. The method for predicting the entire process of rock shear failure based on acoustic emission monitoring according to claim 1, characterized in that: The principle underlying the calculation of the shear component contribution for each acoustic emission event is as follows: The three-dimensional coordinates of each acoustic emission event point are calculated based on the Geiger algorithm. The moment tensor is: in, Represents the moment tensor. This represents the moment of a couple acting in a plane perpendicular to the x-axis. Indicates action on the vertical Moment of couple in the axial plane Indicates action on the vertical Moment of couple in the axial plane express Pure shear couple moment in the plane express Pure shear couple moment in the plane express Pure shear couple moment in the plane; Decomposing the moment tensor into its isotropic and partial tensor components, the moment tensor is expressed as: in, Indicates the isotropic part, Indicates the partial tensor component; The partial tensor part is: in, Represents the trace of the moment tensor. Represents the identity matrix; Based on eigenvalue decomposition, the partial tensor part is decomposed into a double-couple part and a compensated linear vector dipole part: in, Indicates the double couple part. Represents the compensating linear vector dipole component; The formula for calculating the contribution of the shear component is: in, Indicates the contribution of shear components. This represents the Frobenius norm.
5. The method for predicting the entire process of rock shear failure based on acoustic emission monitoring according to claim 1, characterized in that: The principle behind dividing the rock to be tested into several grids in step 3 is as follows: Based on the deployment range of monitoring points, the three-dimensional spatial boundary of the rock to be detected is determined, and then the boundaries of the rock in the x, y, and z directions are determined. The region within the boundary is then uniformly divided into several areas of size [missing information]. Grid, in which, Indicates the grid size. This represents the length of the grid in the x-direction. This indicates the length of the grid in the y-direction. This represents the length of the grid in the z-direction; the grid size is determined based on the positioning accuracy and monitoring requirements of acoustic emission events, using an expert scoring method.
6. The method for predicting the entire process of rock shear failure based on acoustic emission monitoring according to claim 4, characterized in that: The principle underlying the generation of the cumulative energy change rate of the mesh in each time window in step 4 is as follows: The formula used to calculate the cumulative energy of the grid over each time window is: in, Indicates the first Within the first time window The cumulative energy of each grid, Indicates the first The number of acoustic emission events within a time window. Indicates the index of the time window. Indicates the first Within the first time window The acoustic emission event affects the first Shear energy contribution of each grid; The formula used to calculate the rate of change of cumulative energy is: in, Indicates the first Within the first time window The rate of cumulative energy change in each grid, Indicates the first The first acoustic emission event within the time window affects the first... The shear energy contribution of the first grid, i.e., the first grid at the time of the first acoustic emission event. The cumulative energy of each grid, Indicates the first The first acoustic emission event occurred within a certain time window, and the first... The time interval between infrasound emission events.
7. The method for predicting the entire process of rock shear failure based on acoustic emission monitoring according to claim 6, characterized in that: The principle behind establishing the mapping relationship between the cumulative energy change rate of three consecutive time windows and the cumulative energy change rate of the next adjacent time window in step 4 is as follows: A deep learning model is constructed by dividing the historical monitoring data into several time windows according to the monitoring time. Each group of four consecutive time windows is used as a set, and the cumulative energy change rate of each time window is calculated. The cumulative energy change rates of the first three time windows are used as inputs, and the cumulative energy change rate of the fourth time window is used as a label to train the change prediction model.
8. The method for predicting the entire process of rock shear failure based on acoustic emission monitoring according to claim 7, characterized in that: The principle underlying the determination of whether shearing failure occurs in the next time window in step 5 is as follows: The cumulative energy change rate of the mesh, collected and calculated over three consecutive time windows, is input into the change prediction model to obtain the cumulative energy change rate for the next time window. If the cumulative energy change rate of the next time window is greater than or equal to the failure rate threshold, the mesh will undergo shear failure in the next time window; otherwise, shear failure will not occur.
9. A rock shear failure prediction system based on acoustic emission monitoring, characterized in that: The system is used to implement the rock shear failure prediction method based on acoustic emission monitoring as described in any one of claims 1-8, specifically including: The data acquisition and optimization module is used to mark several monitoring points on the surface of the rock to be tested and determine their three-dimensional coordinates. It simultaneously acquires acoustic emission signals at each monitoring point, analyzes the acoustic emission signals of all monitoring points, and combines the acoustic emission signals of all monitoring points to determine the full waveform data of each acoustic emission event. The shearing calculation module is used to obtain the arrival time of the P wave at different monitoring points based on the full waveform data of the acoustic emission event, and determine the three-dimensional coordinates of the occurrence point of each acoustic emission event by combining the three-dimensional coordinates of the corresponding monitoring points; it performs moment tensor inversion based on the full waveform data of each acoustic emission event, and decomposes the moment tensor to calculate the shearing component contribution of each acoustic emission event. The energy contribution calculation module is used to divide the rock to be detected into several grids, calculate the energy influence weight of each acoustic emission event point on each grid, and then calculate its shear energy contribution to each grid. The energy change calculation module is used to divide the monitoring time of acoustic emission events into several time windows of equal length. For each grid, the cumulative energy of each time window is generated based on the shear energy contribution of different acoustic emission events, and then the cumulative energy change rate of the grid in each time window is generated. The mapping relationship between the cumulative energy change rate of three consecutive time windows and the cumulative energy change rate of the next adjacent time window is established. The comprehensive judgment module is used to determine the rate of cumulative energy change when rock undergoes shear failure, which is set as the failure rate threshold. Based on the rate of cumulative energy change of three consecutive time windows, the rate of cumulative energy change of the next time window is obtained, and combined with the failure rate threshold, it is determined whether shear failure occurs in the next time window.