Automatic generation method of tunneling rock burst data analysis report

Abstract

The invention relates to the technical field of geological engineering, in particular to an automatic generation method of a tunneling rock burst data analysis report. An optical fiber acceleration sensor group is arranged in a tunnel to monitor a stress release event in a rock stratum in real time; the occurrence position of a stress release event is calculated by using a multi-difference positioning equation, and accurate space information is provided for subsequent rockburst risk assessment; time domain and frequency domain analysis is carried out on the monitored stress release event, key feature parameters are extracted, analysis feature vectors are constructed, and a scientific basis is provided for rockburst risk assessment; the potential hazard degree of the rockburst event is evaluated by calculating a clustering energy index; according to the method, the fitting surface equation of the discrete event occurrence position coordinates in the micro-seismic event clustering is established, the rockburst early warning line is solved and displayed through the three-dimensional model, the rockburst risk area is visually presented, and a visual reference is provided for constructors.

Description

Technical Field

[0001] The invention relates to the technical field of geological engineering, and in particular to a method for automatically generating a tunnel excavation rock burst data analysis report. Background Art

[0002] Rockburst, also known as rock burst, is a sudden destructive phenomenon that occurs in open rock mass in deep underground mining or areas with high tectonic stress. The cause of rockburst is the sudden and violent release of strain energy accumulated in open rock mass, causing brittle fracture of the rock mass like an explosion. When the high elastic strain energy accumulated in the rock mass is greater than the energy consumed by rock destruction, it will destroy the balance of the rock mass structure, and the excess energy will cause the rock to burst, causing rock fragments to peel off and collapse from the rock mass.

[0003] Rockburst is a common geological disaster in tunnel excavation projects. It can cause serious damage to tunnel structures and construction equipment, and even threaten the lives of construction workers. In order to effectively deal with rockburst, the engineering and academic communities have been exploring and studying rockburst prediction, monitoring and prevention measures.

[0004] However, most of the existing rockburst data analysis methods for tunneling rely on manual operation and judgment, which has many shortcomings. First, manual data analysis is not only time-consuming and laborious, but also easily affected by personal experience and subjective judgment, resulting in inaccurate and inconsistent analysis results. Secondly, with the continuous advancement of tunneling projects, the amount of rockburst data generated is becoming increasingly large, and manual analysis can no longer meet the timeliness and accuracy requirements of data processing. In addition, the existing rockburst analysis software or platform still has some limitations in functionality and ease of use, such as insufficient flexibility in data import and processing, a single presentation method for analysis results, and a lack of user customization and scalability.

[0005] In view of the above problems, it is necessary to propose an automatic generation method for tunnel excavation rockburst data analysis report. Summary of the invention

[0006] The purpose of the present invention is to solve the problems existing in the background technology and to propose a method for automatically generating a tunnel excavation rock burst data analysis report.

[0007] The purpose of the present invention can be achieved through the following technical solutions:

[0008] Step 1: Record stress release events;

[0009] Taking the tunnel line starting pile number K0 as the starting point, a sensor group is arranged at every preset monitoring distance along the tunnel forward direction.

[0010] The sensor group is arranged on the same tunnel cross section and includes at least two optical fiber acceleration sensors respectively located at the arch waist positions on both sides.

[0011] The optical fiber acceleration sensor is placed in a steel sleeve and screwed into a whole with the sleeve by screws. The sleeve is placed in a borehole at a preset distance from the rock mass and coupled with the surrounding rock by an anchoring agent to form a whole.

[0012] All the fiber optic acceleration sensors arranged in the tunnel are numbered with a number symbol i, i=1, 2, 3, ..., n; n is the total number of the fiber optic acceleration sensors.

[0013] A three-dimensional coordinate system is established with the starting pile number K0 of the monitored tunnel line as the origin o, the horizontal plane as the XoY plane, and the vertical direction as the Z axis to obtain the coordinates of each fiber optic acceleration sensor in the three-dimensional coordinate system (xi, yi, zi) = (x1, y1, z1), (x2, y2, z2), ..., (xn, yn, zn).

[0014] Each fiber optic accelerometer indirectly monitors stress release events occurring in nearby rock formations by recording the waveform of the received optical signal.

[0015] After each preset window time T, each fiber optic acceleration sensor i converts the recorded optical signal waveform into a time domain signal Si(t), where 0≤t≤T. The time domain signal reflects the change of the vibration acceleration detected by the sensor over time.

[0016] Step 2: Time domain analysis:

[0017] The maximum amplitude MaxSi(t) and the minimum amplitude MinSi(t) in the time domain signal Si(t) are obtained. The generation time of the maximum amplitude MaxSi(t) is obtained and recorded as the event characteristic time ti of the optical fiber acceleration sensor i.

[0018] By formula Calculate the amplitude fluctuation parameter p1i and the root mean square parameter p2i of the time domain signal Si(t).

[0019] After completing the domain analysis, proceed to step 3.

[0020] Step 3: Frequency domain analysis:

[0021] First, the time domain signal Si(t) is transformed from the time domain to the frequency domain through Fourier transform. The transformation formula is:

[0022] Wherein Si(f) is the frequency domain signal obtained after transformation, f is the frequency; wherein j is the imaginary unit and j×j=-1.

[0023] Then, a proportional signal analysis is performed to obtain the maximum value of the frequency domain signal e0i=MaxSi(f), and the maximum value of the frequency domain signal e1i at the first characteristic frequency f1Min to f1Max and the maximum value of the frequency domain signal e2i at the second characteristic frequency f2Min to f2Max, that is:

[0024]

[0025] By formula Calculate the proportional anomaly factor q1i.

[0026] Finally, perform signal analysis at the differential scale to obtain the maximum value of the first-order derivative of the frequency domain signal Si(f) with respect to frequency.

[0027] Generate the analysis feature vector (p1i, p2i, q1i, q2i) of the optical fiber acceleration sensor i, then end the frequency domain analysis and proceed to step 4.

[0028] Step 4: Compare analysis results and output signals;

[0029] By weighted Euclidean distance formula

[0030] Calculate the analytical eigenvector (p1i, p2i, q1i, q2i) and the reference eigenvector of the fiber optic acceleration sensor i The weighted Euclidean distance Wed(i) is calculated, where k1, k2, k3 and k4 are all preset weight factors.

[0031] When the weighted Euclidean distance Wed(i) is greater than the minimum threshold Wed_Min, it is determined that a stress release event occurs. Subsequently, an event signal is generated and the fiber optic acceleration sensor number symbol i, event feature time ti and sensor coordinates (xi, yi, zi) are output. At the same time, if an event signal is generated, the value of the output state symbol signal(ti, i) is 1, otherwise the value of the output state symbol signal(ti, i) is 0.

[0032] Step 5: Energy analysis and rockburst risk assessment;

[0033] S501, stress release event positioning;

[0034] Every time a window time T passes, a status symbol detection is performed. If a stress release event is detected within the previous window time T, that is, the values ​​of all status symbols output by each fiber optic acceleration sensor are not all 0, the coordinate positioning operation is entered;

[0035] Coordinate positioning operation:

[0036] Assuming that the coordinates of the occurrence position of the stress release event are (x, y, z), the multi-difference positioning equation is established: Where εi is the actual arrival time tolerance error of each fiber optic acceleration sensor i, and ti correct is the actual arrival time after error correction; Vi is the preset basic wave velocity; ri is the arrival time residual, that is, the difference between the theoretical arrival time and the actual arrival time.

[0037] The multi-difference positioning equation is traversed over all fiber optic acceleration sensor serial numbers i to obtain the arrival time residual ri corresponding to each fiber optic acceleration sensor i.

[0038] Set the constraints of the multi-difference positioning equation: Under the constraint conditions, x, y and z are solved to obtain the occurrence position coordinates (x, y, z) of the stress release event in each window time T.

[0039] If no stress release event is detected within a certain window time T, that is, the values ​​of all status symbols output by each fiber optic acceleration sensor are 0, the coordinate positioning operation will not be entered, and the position coordinates will not be output.

[0040] S502, location clustering occurs;

[0041] The occurrence position coordinates of the stress release events generated in each window time T are spatially merged, and a distance threshold d_threshold is preset.

[0042] All stress release events whose distance between occurrence locations is less than the threshold d_threshold are classified into the same event cluster group.

[0043] Specifically, a unique identifier group_id is set up for each event clustering group to distinguish different clustering groups; the group_id number starts from 1 and increases in sequence, that is, group_id=1, 2, 3, ..., m; where m represents the total number of event clustering groups.

[0044] Each event cluster group j includes a number of stress release event occurrence position coordinates (x, y, z), and a unique number u is matched to each stress release event occurrence position coordinate in the same event cluster group; u = 1, 2, 3, ..., u (j). Wherein u (j) is the total number of stress release event occurrence position coordinates contained in event cluster group j.

[0045] The coordinates of the event occurrence locations in event cluster group j are marked as (xu, yu, zu) = (x1, y1, z1), (x2, y2, z2), ..., (xu(j), yu(j), zu(j)).

[0046] Define the center coordinates of each event cluster group

[0047] S503, stress release event frequency analysis;

[0048] Evaluate the frequency and time series concentration of stress release events in each event cluster group in the near future. Get the current time t0, for each event cluster group j, through the formula Calculate the frequency index Count(j), where τ is the preset decay base.

[0049] S504, event aftereffect analysis;

[0050] By formula Calculate the clustering energy index E(j) of each event cluster group.

[0051] The cluster energy index corresponding to each event cluster is marked in the three-dimensional space diagram, and the event cluster whose cluster energy index E exceeds the preset threshold is marked as a high-energy rock burst event cluster and highlighted.

[0052] Event clusters whose clustering energy index E is less than or equal to a preset threshold are marked as microseismic event clusters.

[0053] S505, microseismic event prediction;

[0054] Obtain the coordinates of all discrete event occurrence locations in the microseismic event cluster (xu, yu, zu) = (x1, y1, z1), (x2, y2, z2), ..., (xu(j), yu(j), zu(j)).

[0055] The fitting surface equation for all discrete event location coordinate points is established as z=ax+by+c, where a, b and c are preset fitting surface parameters.

[0056] Establish the loss function:

[0057] Solving Systems of Equations Get the fitting surface parameters a, b and c.

[0058] Furthermore, the intersection line between the fitting surface z=ax+by+c and the tunnel surrounding rock surface is solved and recorded as the rock burst warning line.

[0059] The rockburst warning line is highlighted through the three-dimensional model display, and the mileage pile number where the rockburst warning line is located is output.

[0060] Step 6: Generate rockburst warning and analysis report;

[0061] Use natural language processing technology combined with chart generation algorithms to automatically combine data analysis results with preset report templates and convert them into natural language, automatically fill in analysis results, and generate standardized rockburst data analysis reports.

[0062] The report content includes but is not limited to: event clustering three-dimensional model, three-dimensional model of rock burst warning line, and mileage pile number where the rock burst warning line is located.

[0063] Compared with the prior art, the present invention has the following beneficial effects:

[0064] (1) The present invention can monitor stress release events in the rock formation in real time and realize instant perception of rock bursts by arranging a group of fiber optic acceleration sensors in the tunnel. By using the multi-difference positioning equation, the location of the stress release event can be accurately calculated to provide accurate spatial information for subsequent rock burst risk assessment; the monitored stress release events are analyzed in the time domain and frequency domain, key characteristic parameters are extracted, and analysis characteristic vectors are constructed to provide a scientific basis for rock burst risk assessment. By comparing the weighted Euclidean distance formula with the benchmark characteristic vector, it is possible to accurately determine whether the stress release event poses a rock burst threat and output a corresponding warning signal. Comprehensive analysis and intelligent warning of rock bursts are achieved;

[0065] (2) The present invention spatially merges the locations of stress release events to form event clustering groups, which is convenient for analyzing the distribution patterns and trends of rock burst events. By calculating the clustering energy index, the potential harm degree of rock burst events can be evaluated, providing a basis for formulating targeted prevention and control measures;

[0066] (3) The present invention establishes a fitting surface equation for the coordinates of the locations of discrete events in microseismic event clustering, and solves it to obtain the rockburst warning line, thereby achieving accurate prediction of potential rockburst areas. The rockburst warning line is displayed through a three-dimensional model, which intuitively presents the rockburst risk area and provides an intuitive visual reference for construction personnel. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings:

[0068] Figure 1 is a flow chart of the method of the present invention;

[0069] Figure 2 It is a schematic diagram of the arrangement of the sensor group of the present invention;

[0070] Figure 3 It is a schematic diagram of the time domain signal of the optical fiber acceleration sensor;

[0071] Figure 4 This is a schematic diagram of event clustering results;

[0072] Figure 5Schematic diagram of the fitting surface and rockburst warning line. DETAILED DESCRIPTION

[0073] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0074] See also Figure 1 As shown, a method for automatically generating a tunnel excavation rock burst data analysis report comprises the following steps:

[0075] Step 1: Record stress release events;

[0076] See also Figure 2 As shown, starting from the tunnel line starting pile number K0, a sensor group is arranged at every preset monitoring distance along the tunnel forward direction.

[0077] The sensor group is arranged on the same tunnel cross section and includes at least two optical fiber acceleration sensors respectively located at the arch waist positions on both sides.

[0078] The optical fiber acceleration sensor is placed in a steel sleeve and screwed into a whole with the sleeve by screws. The sleeve is placed in a borehole at a preset distance from the rock mass and coupled with the surrounding rock by an anchoring agent to form a whole.

[0079] All the fiber optic acceleration sensors arranged in the tunnel are numbered with a number symbol i, i=1, 2, 3, ..., n; n is the total number of the fiber optic acceleration sensors.

[0080] A three-dimensional coordinate system is established with the starting pile number K0 of the monitored tunnel line as the origin o, the horizontal plane as the XoY plane, and the vertical direction as the Z axis to obtain the coordinates of each fiber optic acceleration sensor in the three-dimensional coordinate system (xi, yi, zi) = (x1, y1, z1), (x2, y2, z2), ..., (xn, yn, zn).

[0081] Each fiber optic accelerometer indirectly monitors stress release events occurring in nearby rock formations by recording the waveform of the received optical signal.

[0082] It should be noted that stress release events, such as rock bursts, are not only accompanied by rock fractures during their development, but also trigger seismic waves. When these seismic waves pass through the rock layer and act on the optical fiber sensor, the optical fiber will produce tiny strains, which will cause slight changes in the frequency and phase of the optical signal transmitted inside it. By recording the optical signal waveform, the frequency, intensity and position of the vibration point can be indirectly recorded, thereby realizing the monitoring and analysis of rock stress release events.

[0083] See also Figure 3 As shown, after each preset window time T, each optical fiber acceleration sensor i converts the recorded optical signal waveform into a time domain signal Si(t), where 0≤t≤T. The time domain signal reflects the change of the vibration acceleration detected by the sensor over time.

[0084] Step 2: Time domain analysis:

[0085] The maximum amplitude MaxSi(t) and the minimum amplitude MinSi(t) in the time domain signal Si(t) are obtained. The generation time of the maximum amplitude MaxSi(t) is obtained and recorded as the event characteristic time ti of the optical fiber acceleration sensor i.

[0086] By formula Calculate the amplitude fluctuation parameter p1i and the root mean square parameter p2i of the time domain signal Si(t).

[0087] After completing the domain analysis, proceed to step 3.

[0088] Step 3: Frequency domain analysis:

[0089] First, the time domain signal Si(t) is transformed from the time domain to the frequency domain through Fourier transform. The transformation formula is:

[0090] Wherein Si(f) is the frequency domain signal obtained after transformation, f is the frequency; wherein j is the imaginary unit and j×j=-1.

[0091] Then, a proportional signal analysis is performed to obtain the maximum value of the frequency domain signal e0i=MaxSi(f), and the maximum value of the frequency domain signal e1i at the first characteristic frequency f1Min to f1Max and the maximum value of the frequency domain signal e2i at the second characteristic frequency f2Min to f2Max, that is:

[0092]

[0093] By formula Calculate the proportional anomaly factor q1i.

[0094] Finally, perform signal analysis at the differential scale to obtain the maximum value of the first-order derivative of the frequency domain signal Si(f) with respect to frequency.

[0095] Generate the analysis feature vector (p1i, p2i, q1i, q2i) of the optical fiber acceleration sensor i, then end the frequency domain analysis and proceed to step 4.

[0096] Step 4: Compare analysis results and output signals;

[0097] By weighted Euclidean distance formula

[0098] Calculate the analytical eigenvector (p1i, p2i, q1i, q2i) and the reference eigenvector of the fiber optic acceleration sensor i The weighted Euclidean distance Wed(i) is calculated, where k1, k2, k3 and k4 are all preset weight factors.

[0099] When the weighted Euclidean distance Wed(i) is greater than the minimum threshold Wed_Min, it is determined that a stress release event occurs. Subsequently, an event signal is generated and the fiber optic acceleration sensor number symbol i, event feature time ti and sensor coordinates (xi, yi, zi) are output. At the same time, if an event signal is generated, the value of the output state symbol signal(ti, i) is 1, otherwise the value of the output state symbol signal(ti, i) is 0.

[0100] Step 5: Energy analysis and rockburst risk assessment;

[0101] S501, stress release event positioning;

[0102] Every time a window time T passes, a status symbol detection is performed. If a stress release event is detected within the previous window time T, that is, the values ​​of all status symbols output by each fiber optic acceleration sensor are not all 0, the coordinate positioning operation is entered;

[0103] Coordinate positioning operation:

[0104] Assuming that the coordinates of the occurrence position of the stress release event are (x, y, z), the multi-difference positioning equation is established: Where εi is the actual arrival time tolerance error of each fiber optic acceleration sensor i, and ti correct is the actual arrival time after error correction; Vi is the preset basic wave velocity; ri is the arrival time residual, that is, the difference between the theoretical arrival time and the actual arrival time.

[0105] The multi-difference positioning equation is traversed over all fiber optic acceleration sensor serial numbers i to obtain the arrival time residual ri corresponding to each fiber optic acceleration sensor i.

[0106] Set the constraints of the multi-difference positioning equation: Under the constraint conditions, x, y and z are solved to obtain the occurrence position coordinates (x, y, z) of the stress release event in each window time T.

[0107] If no stress release event is detected within a certain window time T, that is, the values ​​of all status symbols output by each fiber optic acceleration sensor are 0, the coordinate positioning operation will not be entered, and the position coordinates will not be output.

[0108] S502, location clustering occurs;

[0109] The occurrence position coordinates of the stress release events generated in each window time T are spatially merged, and a distance threshold d_threshold is preset.

[0110] All stress release events whose distance between occurrence locations is less than the threshold d_threshold are classified into the same event cluster group.

[0111] Specifically, a unique identifier group_id is set up for each event clustering group to distinguish different clustering groups; the group_id number starts from 1 and increases in sequence, that is, group_id=1, 2, 3, ..., m; where m represents the total number of event clustering groups.

[0112] Each event cluster group j includes a number of stress release event occurrence position coordinates (x, y, z), and a unique number u is matched to each stress release event occurrence position coordinate in the same event cluster group; u = 1, 2, 3, ..., u (j). Wherein u (j) is the total number of stress release event occurrence position coordinates contained in event cluster group j.

[0113] The coordinates of the event occurrence locations in event cluster group j are marked as (xu, yu, zu) = (x1, y1, z1), (x2, y2, z2), ..., (xu(j), yu(j), zu(j)).

[0114] Define the center coordinates of each event cluster group

[0115] It should be noted that each event cluster group represents a group of stress release events that occur continuously in time and are close in space. The central coordinates of these event cluster groups provide the geometric center positions of these events in spatial distribution.

[0116] S503, stress release event frequency analysis;

[0117] Evaluate the frequency and time series concentration of stress release events in each event cluster group in the near future. Get the current time t0, for each event cluster group j, through the formula Calculate the frequency index Count(j), where τ is the preset decay base.

[0118] It should be noted that the frequency index is proportional to the frequency, concentration and time proximity of stress release events. The larger the value of the frequency index, the more stress release events have been recorded in event cluster group j in the recent period.

[0119] S504, event aftereffect analysis;

[0120] By formula Calculate the clustering energy index E(j) of each event cluster group.

[0121] See also Figure 4 As shown, the cluster energy index corresponding to each event cluster is marked in the three-dimensional space diagram, and the event cluster whose cluster energy index E exceeds a preset threshold is marked as a high-energy rock burst event cluster and highlighted.

[0122] Event clusters whose clustering energy index E is less than or equal to a preset threshold are marked as microseismic event clusters.

[0123] It should be noted that microseismic events in tunnels are often accompanied by crack expansion, joint development and stress concentration. Before a rockburst occurs, microseismic events will gradually increase, accompanied by large-energy microseismic events. The spatial distribution of microseismic events has an obvious correspondence with the location of the rockburst or the stress concentration area, and the recording of microseismic events can visualize the stress transfer path or concentration location, which has an early warning effect on rockburst.

[0124] S505, microseismic event prediction;

[0125] Obtain the coordinates of all discrete event occurrence locations in the microseismic event cluster (xu, yu, zu) = (x1, y1, z1), (x2, y2, z2), ..., (xu(j), yu(j), zu(j)).

[0126] The fitting surface equation for all discrete event location coordinate points is established as z=ax+by+c, where a, b and c are preset fitting surface parameters.

[0127] Establish the loss function:

[0128] Solving Systems of Equations Get the fitting surface parameters a, b and c.

[0129] Furthermore, the intersection line between the fitting surface z=ax+by+c and the tunnel surrounding rock surface is solved and recorded as the rock burst warning line.

[0130] See also Figure 5 As shown, the rockburst warning line is highlighted through the three-dimensional model display, and the mileage pile number where the rockburst warning line is located is output.

[0131] Step 6: Generate rockburst warning and analysis report;

[0132] Use natural language processing technology combined with chart generation algorithms to automatically combine data analysis results with preset report templates and convert them into natural language, automatically fill in analysis results, and generate standardized rockburst data analysis reports.

[0133] The report content includes but is not limited to: event clustering three-dimensional model, three-dimensional model of rock burst warning line, and mileage pile number where the rock burst warning line is located.

[0134] It should be understood that the terms “include” and “comprising” used in the specification and claims of the present disclosure indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections thereof.

[0135] It should also be understood that the terms used in this disclosure are only for the purpose of describing specific embodiments and are not intended to limit the disclosure. As used in this disclosure and the claims, the singular forms "a", "an", and "the" are intended to include the plural forms unless the context clearly indicates otherwise. It should also be further understood that the term "and / or" used in this disclosure and the claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes these combinations;

[0136] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for automatically generating a tunnel excavation rock burst data analysis report, characterized in that: The following steps are involved: Step 1: stress release event recording; arrange a fiber optic acceleration sensor group in the tunnel; monitor the stress release event generated in the nearby rock formation by recording the waveform of the received optical signal through the fiber optic acceleration sensor to obtain a time domain signal Si(t); Step 2: Time domain analysis: convert the optical signal waveform recorded by each optical fiber acceleration sensor into a time domain signal, and obtain the maximum amplitude, minimum amplitude and generation time of the time domain signal; calculate the amplitude fluctuation parameter and root mean square parameter of the time domain signal; Step 3, frequency domain analysis: convert the time domain signal into a frequency domain signal through Fourier transform; perform proportional scale and differential scale signal analysis on the frequency domain signal to calculate the proportional anomaly factor and generate the analysis feature vector of the optical fiber acceleration sensor; Step 4: Compare the analysis results and output the signal; calculate the weighted Euclidean distance between the analysis feature vector and the reference feature vector by using the weighted Euclidean distance formula; when the weighted Euclidean distance is greater than the minimum threshold, determine that a stress release event has occurred, and output the corresponding sensor number, event feature time, sensor coordinates and status symbol; Step 5: Energy analysis and rockburst risk assessment; S501, stress release event location; S502, occurrence location clustering; S503, stress release event frequency analysis; S504, event aftereffect analysis; S505, microseismic event prediction; Step 6: Generate rockburst warning and analysis report; Use natural language processing technology combined with chart generation algorithms to automatically combine data analysis results with preset report templates and convert them into natural language, automatically fill in analysis results, and generate standardized rockburst data analysis reports; report content includes but is not limited to: event clustering 3D model, 3D model of rockburst warning line, and mileage pile number where the rockburst warning line is located.

2. The method for automatically generating a tunnel excavation rockburst data analysis report according to claim 1, characterized in that: The arrangement method and monitoring process of the optical fiber acceleration sensor group described in step 1 are specifically as follows: Taking the tunnel line starting stake number K0 as the starting point, a sensor group is arranged at every preset monitoring distance along the tunnel forward direction; The sensor group is arranged on the same tunnel cross section and includes at least two optical fiber acceleration sensors respectively located at the arch waist positions on both sides; The optical fiber acceleration sensor is placed in a steel sleeve and screwed into a whole with the sleeve by screws. The sleeve is placed in a borehole at a preset distance from the rock mass and coupled with the surrounding rock by an anchoring agent to form a whole. All the optical fiber acceleration sensors arranged in the tunnel are numbered, and the number symbol is i, i=1, 2, 3, ..., n; n is the total number of fiber optic acceleration sensors; A three-dimensional coordinate system is established with the starting stake number K0 of the monitored tunnel line as the origin o, the horizontal plane as the XoY plane, and the vertical direction as the Z axis, and the coordinates of each fiber optic acceleration sensor in the three-dimensional coordinate system are obtained (xi, yi, zi) = (x1, y1, z1), (x2, y2, z2), ..., (xn, yn, zn); Each fiber optic acceleration sensor indirectly monitors stress release events generated in nearby rock formations by recording the waveform of the received optical signal; After each preset window time T, each fiber optic acceleration sensor i converts the recorded optical signal waveform into a time domain signal Si(t), where 0≤t≤T. The time domain signal reflects the change of the vibration acceleration detected by the sensor over time.

3. The method for automatically generating a tunnel excavation rockburst data analysis report according to claim 1, characterized in that: The specific process of time domain analysis described in step 2 is: Obtain the maximum amplitude MaxSi(t) and the minimum amplitude MinSi(t) in the time domain signal Si(t); obtain the generation time of the maximum amplitude MaxSi(t), which is recorded as the event characteristic time ti of the optical fiber acceleration sensor i; By formula Calculate the amplitude fluctuation parameter p1i and the root mean square parameter p2i of the time domain signal Si(t).

4. The method for automatically generating a tunnel excavation rockburst data analysis report according to claim 1, characterized in that: The specific process of frequency domain analysis described in step 3 is: First, the time domain signal Si(t) is transformed from the time domain to the frequency domain through Fourier transform. The transformation formula is: Where Si(f) is the frequency domain signal obtained after transformation, f is the frequency; where j is the imaginary unit and j×j=-1; Then, a proportional signal analysis is performed to obtain the maximum value of the frequency domain signal e0i=MaxSi(f), and the maximum value of the frequency domain signal e1i at the first characteristic frequency f1Min to f1Max and the maximum value of the frequency domain signal e2i at the second characteristic frequency f2Min to f2Max, that is: By formula Calculate the proportional anomaly factor q1i; Finally, perform signal analysis at the differential scale to obtain the maximum value of the first-order derivative of the frequency domain signal Si(f) with respect to frequency. Generate the analysis feature vector (p1i, p2i, q1i, q2i) of the optical fiber acceleration sensor i, and then end the frequency domain analysis.

5. The method for automatically generating a tunnel excavation rockburst data analysis report according to claim 1, characterized in that: The specific process of weighted Euclidean distance calculation described in step 4 is: By weighted Euclidean distance formula Calculate the analytical eigenvector (p1i, p2i, q1i, q2i) and the reference eigenvector of the fiber optic acceleration sensor i The weighted Euclidean distance Wed(i) of , where k1, k2, k3 and k4 are all preset weight factors; When the weighted Euclidean distance Wed(i) is greater than the minimum threshold Wed_Min, it is determined that a stress release event has occurred; subsequently, an event signal is generated and the fiber optic acceleration sensor number symbol i, event characteristic time ti and sensor coordinates (xi, yi, zi) are output; at the same time, if an event signal is generated, the value of the output status symbol signal(ti, i) is 1, otherwise the value of the output status symbol signal(ti, i) is 0.

6. The method for automatically generating a tunnel excavation rockburst data analysis report according to claim 1, characterized in that: The specific process of the coordinate positioning operation described in S501 is: Assuming that the coordinates of the occurrence position of the stress release event are (x, y, z), the multi-difference positioning equation is established: Where εi is the actual arrival time tolerance error of each fiber optic acceleration sensor i, and ti correct is the actual arrival time after error correction; Where Vi is the preset basic wave speed; where ri is the arrival time residual, that is, the difference between the theoretical arrival time and the actual arrival time; The multi-difference positioning equation is traversed through all fiber optic acceleration sensor number symbols i to obtain the arrival time residual ri corresponding to each fiber optic acceleration sensor i; Set the constraints of the multi-difference positioning equation: Solve x, y and z under the constraint conditions to obtain the occurrence position coordinates (x, y, z) of the stress release event in each window time T; If no stress release event is detected within a certain window time T, that is, the values ​​of all status symbols output by each fiber optic acceleration sensor are 0, the coordinate positioning operation will not be entered, and the position coordinates will not be output.

7. The method for automatically generating a tunnel excavation rockburst data analysis report according to claim 1, characterized in that: The specific process of location clustering described in S502 is as follows: The occurrence position coordinates of the stress release events generated in each window time T are spatially merged, and a distance threshold d_threshold is preset; All stress release events whose distance between occurrence locations is less than the threshold d_threshold are classified into the same event cluster group; Specifically, a unique identifier group_id is set for each event clustering group to distinguish different clustering groups; the group_id number starts from 1 and increases in sequence, that is, group_id = 1, 2, 3, ..., m; where m represents the total number of event clustering groups; Each event cluster group j includes a number of stress release event occurrence position coordinates (x, y, z), and a unique number u is matched to each stress release event occurrence position coordinate in the same event cluster group; u = 1, 2, 3, ..., u(j); wherein u(j) is the total number of stress release event occurrence position coordinates contained in event cluster group j; The coordinates of the event occurrence locations in event cluster group j are marked as (xu, yu, zu) = (x1, y1, z1), (x2, y2, z2), ..., (xu(j), yu(j), zu(j)); Define the center coordinates of each event cluster group 8. The method for automatically generating a tunnel excavation rockburst data analysis report according to claim 1, characterized in that: The specific process of stress release event frequency analysis described in S503 is: Get the current time t0, for each event cluster group j, through the formula Calculate the frequency index Count(j), where τ is the preset decay base.

9. The method for automatically generating a tunnel excavation rockburst data analysis report according to claim 1, characterized in that: The specific process of the event aftereffect analysis described in S504 is as follows: By formula Calculate the clustering energy index E(j) of each event cluster group; The cluster energy index corresponding to each event cluster is marked in the three-dimensional space diagram, and the event cluster whose cluster energy index E exceeds the preset threshold is marked as a high-energy rock burst event cluster and highlighted; Event clusters whose clustering energy index E is less than or equal to a preset threshold are marked as microseismic event clusters.

10. The method for automatically generating a tunnel excavation rockburst data analysis report according to claim 1, characterized in that: The specific process of microseismic event prediction described in S505 is: Obtain the coordinates of all discrete event occurrence locations in the microseismic event cluster (xu, yu, zu) = (x1, y1, z1), (x2, y2, z2), ..., (xu(j), yu(j), zu(j)); The fitting surface equation for all discrete event location coordinate points is established as z=ax+by+c; where a, b and c are preset fitting surface parameters; Establish the loss function: Solving Systems of Equations Get the fitting surface parameters a, b and c; The intersection line between the fitting surface z=ax+by+c and the tunnel surrounding rock surface is solved and recorded as the rock burst warning line. The rockburst warning line is highlighted through the three-dimensional model display, and the mileage pile number where the rockburst warning line is located is output.

Images