Coal rock mass fracture instability multivariate signal fusion analysis method

Abstract

The invention discloses a coal and rock mass fracture instability multivariate signal fusion analysis method, and belongs to coal and rock mass analysis. The method comprises the following steps: carrying out time-frequency domain processing on an acoustic emission signal to obtain an accumulated ringing characteristic value # imgabs0 # and an accumulated b characteristic value # imgabs1 #; a main control frequency characteristic value # imgabs2 # and an amplitude characteristic value # imgabs3 # are accumulated; a spatial characteristic value # imgabs4 # of acoustic emission; fitting a fitting function # imgabs7 # from the characteristic value # imgabs5 # to the characteristic value # imgabs6 # respectively; according to the fitting function # imgabs8 #, calculating a correlation coefficient matrix R between the characteristic values # imgabs9 # to # imgabs10 #; according to the matrix R, determining a weight coefficient # imgabs11 # of each characteristic value; a first-order derivative function # imgabs13 # and a second-order derivative function # imgabs14 # of the fitting function # imgabs12 # are obtained, and according to the first-order derivative function # imgabs15 # and the second-order derivative function # imgabs16 #, a demarcation time point # imgabs17 # in the coal and rock mass damage process is determined; and the boundary time point # imgabs19 # is corrected by using the weight coefficient # imgabs18 #, and a corrected boundary time point # imgabs20 # is obtained. In order to solve the problems that an existing coal and rock mass fracture instability judgment index is single, and coal and rock mass fracture instability early warning is low in single index precision, the early warning precision is improved through multi-element signal fusion.

Description

Technical Field

[0001] This application relates to the analysis of coal and rock masses, and more specifically, to a multi-signal fusion analysis method for the fracture instability of coal and rock masses. Background Art

[0002] As a non-destructive dynamic monitoring technique, acoustic emission technology has been widely applied in recent years in multiple fields such as mineral resource development, slope protection, railway and highway construction, and tunnel stability. Based on the principle that strain energy is released during the process of coal and rock masses in coal mine carbon storage spaces being stressed and deformed and fractured, and is propagated outward in the form of stress waves, this technology can invert the fracture information and the dynamic evolution of internal micro-fractures during the fracture process of coal and rock masses, thus effectively understanding the fracture mechanism and failure precursors of coal and rock masses, etc., providing important technical support for the instability early warning analysis of coal and rock masses.

[0003] The acoustic emission waveforms generated during the failure process of coal and rock masses carry rich information on parameters such as their stress states, internal structures, crack transformations, and physical and mechanical properties. By analyzing the acoustic emission wave spectrum signals and spatial characteristic information, the deformation and failure mechanism and failure precursor information of coal and rock masses can be more comprehensively understood. This is of great significance for the safety monitoring and early warning of underground rock engineering.

[0004] However, there are obvious deficiencies in the prior art in the early warning of coal and rock mass fracture instability. Traditional methods mainly use single time-domain features or frequency-domain features for analysis, lack cumulative effect analysis, have a large degree of arbitrariness in parameter selection, cannot fully reflect the particularity of coal and rock mass fracture failure, the early warning information is inaccurate and untimely, and it is difficult to meet the actual engineering requirements. In addition, the prior art generally ignores the spatial distribution characteristics of acoustic emission signals and cannot effectively capture the spatially non-uniform failure characteristics preferentially caused along fractures.

[0005] For example, the related patent document CN118688301A discloses a method for evaluating the failure of fractured rock masses based on acoustic emission characteristic parameters, including: constructing an acoustic emission system to obtain the acoustic emission signals generated by a rock sample during the stress application process, inputting the acoustic emission signals into the acoustic emission system to obtain the acoustic emission amplitude, acoustic emission magnitude, and acoustic emission ringing count rate; calculating the acoustic emission b-value based on the acoustic emission amplitude and acoustic emission magnitude; calculating the correlation dimension value D2 based on the acoustic emission ringing count rate; during the change process of the correlation dimension value D2, introducing a process parameter sequence T to ensure that the calculation step distances of the acoustic emission b-value and the correlation dimension value D2 are consistent, and when the correlation dimension value D2 is the largest and the acoustic emission b-value is the smallest, finding the failure point. However, on the one hand, this solution only relies on two parameters, the acoustic emission b-value and the correlation dimension D2, for judgment, and the signal feature dimension is too low to comprehensively characterize the complex process of the internal microcrack evolution of coal and rock masses. On the other hand, it completely ignores the influence of the interaction between fluids (such as gas, water, etc.) and coal and rock masses, which is a key factor leading to the instability of coal and rock masses in the actual mine environment. Summary of the Invention

[0006] In view of the fact that the interaction mechanism between fluids and coal and rock masses is often ignored in the early warning of the instability of coal and rock masses in the prior art and the analysis index is single, resulting in low early warning accuracy, the present application provides a multi-signal fusion analysis method for the fracture instability of coal and rock masses, a multi-signal fusion analysis method based on the interaction between fluid and coal and rock masses, which improves the early warning accuracy.

[0007] The present application provides a multi-signal fusion analysis method for the fracture instability of coal and rock masses, including: collecting acoustic emission data during the deformation and failure process of coal and rock masses; the acoustic emission data includes acoustic emission signals and spatial data corresponding to the acoustic emission signals; performing time-frequency domain processing on the acoustic emission signals to obtain the cumulative ringing characteristic value of acoustic emission , the cumulative b characteristic value , the cumulative main control frequency characteristic value of acoustic emission and the cumulative amplitude characteristic value ; extracting the cumulative acoustic emission energy value from the spatial data as the spatial characteristic value of acoustic emission ; extracting the cumulative acoustic emission energy value from the spatial data as the spatial characteristic value of acoustic emission ; respectively fitting the fitting functions of the characteristic values to the characteristic values ; according to the fitting function , calculating the correlation coefficient matrix R between the characteristic values to ; according to the correlation coefficient matrix R, determining the weight coefficients of each characteristic value ; obtaining the first derivative function and the second derivative function of the fitting function , and according to the first-order and the second-order derivative function , determine the demarcation time point during the failure process of the coal and rock mass ; use the weight coefficient to correct the demarcation time point to obtain the corrected demarcation time point , and according to the corrected demarcation time point , divide the failure process of the coal and rock mass into a stable period, a development period, an energy storage period, and a failure period; obtain the time range when the coal and rock mass enters the failure period as the early warning signal for the instability precursor of the coal and rock mass.

[0008] Furthermore, the cumulative ringing eigenvalue represents the cumulative value of the ringing count during the acoustic emission process, reflecting the number of microcracks inside the coal and rock mass; the cumulative b eigenvalue represents the cumulative value of the amplitude-frequency distribution slope coefficient of the acoustic emission signal, reflecting the change in the distribution of the crack scale inside the coal and rock mass during the interaction between the fluid and the coal and rock mass; the cumulative dominant frequency eigenvalue represents the time-series cumulative value of the frequency corresponding to the maximum amplitude in the acoustic emission signal spectrum, reflecting the propagation rate of microcracks inside the coal and rock mass; the cumulative amplitude eigenvalue represents the time-series cumulative value of the maximum amplitude value of the acoustic emission signal, reflecting the intensity change of the energy released by microcracks inside the coal and rock mass.

[0009] Furthermore, to obtain the cumulative b eigenvalue , it includes: extracting the absolute energy E of the acoustic emission from the collected acoustic emission data, and the hit count where the absolute energy is greater than E; establishing the relationship between the acoustic emission absolute energy E and the hit count : , where a and b are constants; set the initial and the step size , and use the sliding window algorithm to segment the acoustic emission data; calculate the cumulative variance ratio CVR of the acoustic emission signal within the current window, where CVR represents the ratio of the signal energy variance to the average energy within the window; according to CVR, calculate the optimal window of the current window, , γ is a positive coefficient; according to the optimal window , calculate the acoustic emission b value of each window by the least squares method; accumulate the acoustic emission b values within each window in chronological order to obtain the acoustic emission cumulative b eigenvalue .

[0010] Furthermore, the cumulative dominant frequency eigenvalue and the cumulative amplitude eigenvalue , including: dividing the process of the interaction between fluid and coal-rock mass into an initial adsorption stage, a diffusion penetration stage, and a fracture propagation stage according to acoustic emission data; respectively setting the window lengths of the sliding windows in the initial adsorption stage, the diffusion penetration stage, and the fracture propagation stage to be , and ; where ; according to the window , and , segmenting the acoustic emission wave data in each stage to obtain the acoustic emission time-domain signals within multiple time windows ; performing wavelet transform on the acoustic emission data within each window to obtain wavelet coefficients; calculating the wavelet energy distribution according to the wavelet coefficients, and dividing the wavelet energy distribution into a low-frequency band, a middle-frequency band, and a high-frequency band; performing peak analysis on each frequency band to extract the low-frequency band main control frequency and the corresponding maximum amplitude , the middle-frequency band main control frequency and the corresponding maximum amplitude , the high-frequency band main control frequency and the corresponding maximum amplitude ; where the main control frequency represents the frequency at which the peak of the wavelet energy distribution is located in the corresponding frequency band; accumulating the main control frequencies of each time window to obtain the cumulative main control frequency eigenvalue ; accumulating the maximum amplitudes of each time window to obtain the cumulative amplitude eigenvalue .

[0011] Specifically, the fracture instability of coal-rock mass often has complex characteristics such as non-linear evolution, stage mutation, and being affected by multi-field coupling. The existing technologies usually adopt Fourier transform or fast Fourier transform, but there are defects in the traditional FFT in dealing with the fracture instability analysis of coal-rock mass: the quasi-FFT is restricted by the Heisenberg uncertainty principle of time-frequency resolution, that is, the fixed number of sampling points N determines both the frequency resolution and the time window length . In the interaction between fluid and coal-rock mass, the coal-rock mass will experience multi-stage damage from microscopic adsorption, mesoscopic diffusion to macroscopic fracture propagation, and different stages require different time-frequency resolutions. When the micro-cracks of the coal-rock mass rapidly expand, a higher time resolution is required, while when monitoring the change of matrix adsorption, a higher frequency resolution is required, and the fixed-parameter FFT cannot meet these two requirements at the same time.

[0012] This application proposes an automatic discrimination algorithm based on the acoustic emission event rate and the proportion of band energy, which scientifically divides the interaction between fluid and coal-rock mass into an initial adsorption stage, a diffusion and penetration stage, and a crack propagation stage. For each stage, a differential window length strategy is adopted to solve the contradiction that the traditional fixed window has insufficient resolution in the rapidly changing stage and large noise interference in the stable stage. Feature parameters are extracted from the low-frequency band, medium-frequency band, and high-frequency band respectively for different stages, enabling the system to accurately capture the whole-process microscopic signals from microcrack propagation from the microscopic to the macroscopic level.

[0013] Furthermore, the process of the interaction between fluid and coal-rock mass is divided into an initial adsorption stage, a diffusion and penetration stage, and a crack propagation stage, including: calculating the acoustic emission event rate and the proportion of band energy based on the acoustic emission data; setting the first threshold and the second threshold of the acoustic emission event rate; wherein, ; setting the third threshold , the fourth threshold , and the fifth threshold of the proportion of band energy in the low-frequency band, medium-frequency band, and high-frequency band respectively; when the acoustic emission event rate is less than or equal to and the proportion of band energy in the low-frequency band is greater than ε3, it is judged as the initial adsorption stage; when the acoustic emission event rate is greater than and less than , and the proportion of band energy in the medium-frequency band is greater than , it is judged as the diffusion and penetration stage; when the acoustic emission event rate is greater than , and the proportion of band energy in the high-frequency band is greater than

[0014] , it is judged as the crack propagation stage. Specifically, the acoustic emission event rate actually reflects the number of microcracks generated per unit time. In different stages, the physical mechanisms of microcrack generation are different. Specifically, in the initial adsorption stage: fluid molecules mainly adsorb on the surface of the coal-rock mass, and at this time, fewer microcracks are generated, so the acoustic emission event rate is relatively low (≤ <event rate< ); in the diffusion and penetration stage: fluid molecules begin to diffuse and penetrate into the coal-rock mass, resulting in changes in the internal structure and generating more microcracks, and the event rate increases ( ); in the crack propagation stage: the formed microcracks rapidly expand and connect under the action of the fluid, generating a large number of microcracks, and the event rate increases significantly (>

[0015] In addition, different types of microcracks generate acoustic emission signals with different characteristics, which are manifested as the distribution of energy in different frequency bands in the frequency domain. Low-frequency band energy (initial adsorption stage): mainly corresponding to the minute deformation caused by surface adsorption, with low energy conversion efficiency and relatively low frequency; medium-frequency band energy (diffusion and penetration stage): corresponding to the seepage and diffusion processes of fluids inside coal and rock masses, resulting in medium-scale internal structural changes; high-frequency band energy (fracture propagation stage): corresponding to rapid crack propagation and connection, releasing high-frequency acoustic energy.

[0016] Therefore, this application simultaneously uses "acoustic emission event rate" and "proportion of band energy" as the judgment basis. The acoustic emission event rate reflects the time density of the number of microcracks; the proportion of band energy reflects the spectral characteristics of the types of microcracks.

[0017] Furthermore, the wavelet transform adopts the following formula: , where is the wavelet coefficient, s is the scale parameter, τ is the translation parameter, is the time-domain signal; is the conjugate function of the wavelet basis function; the wavelet transform has better time-frequency localization characteristics compared with the traditional Fourier transform and can simultaneously obtain the time-domain and frequency-domain information of the signal; multi-band analysis realizes the classification and identification of cracks in coal and rock masses of different scales: low frequency mainly corresponds to large-scale structural changes, medium frequency corresponds to mesoscopic crack propagation, and high frequency corresponds to microcrack initialization; by analyzing the variation laws of the main control frequencies and the maximum amplitudes of each frequency band, the internal structural evolution characteristics of the fracture instability of coal and rock masses are accurately characterized.

[0018] Furthermore, the fitting functions from eigenvalue to eigenvalue are respectively fitted, including: , where respectively correspond to five eigenvalues, x is the time variable, representing the duration from the start of the interaction between the fluid and the coal and rock mass to the current time, is the j-th coefficient of the i-th eigenvalue fitting function; is the constant term of the polynomial, representing the initial state value of the eigenvalue; j represents the highest degree of polynomial fitting.

[0019] Furthermore, the weight coefficients of each eigenvalue are determined, including: setting the common effective time interval of the fitting function as the minimum domain; according to the minimum domain, the eigenvalue is calculated at the sampling time points through the fitting function as the sample data; according to the sample data, an acoustic emission eigenvalue sample matrix F is constructed: , where denotes the value of the i-th eigenvalue at the time point , where m is the number of sampling points; according to the acoustic emission eigenvalue sample matrix F, calculate the eigenvalues to and calculate the correlation coefficient matrix R between them. The element in matrix R denotes the correlation coefficient between the eigenvalues and ; calculate the matrix eigenvalues to of the correlation coefficient matrix R, and the corresponding eigenvectors ; calculate the contribution rate and cumulative contribution rate of each matrix eigenvalue , and select the matrix eigenvalues whose cumulative contribution rate is greater than the threshold; according to the selected matrix eigenvalues and the corresponding eigenvectors , construct the factor loading matrix ρ, where ; according to the factor loading matrix ρ, determine the weight coefficients to of the eigenvalues , .

[0020] In particular, the principal component information is extracted through matrix eigenvalue decomposition to remove redundancy and retain the information most sensitive to the state change of the coal and rock mass; the construction of the factor loading matrix realizes the mapping from the original features to the principal component space and optimizes the feature combination; the calculation method of the weight coefficient considers the feature contribution rate and realizes the reasonable quantification of feature importance.

[0021] Furthermore, the expression is as follows: ; where is the correlation coefficient between the eigenvalues and , ; and , and are the average values of and respectively; denotes the value of the eigenvalue at the k-th time point, denotes the value of the eigenvalue at the k-th time point; k represents the serial number of the time point, .

[0022] Furthermore, determine the demarcation time point in the failure process of the coal and rock mass, including: calculating the time point of the first derivative function as the extreme point or stationary point of the fitting function; calculating the second derivative function The time point is used as the inflection point of the fitting function; according to the first derivative function and the second derivative function , as well as the extreme points or stationary points, and the inflection points, the demarcation time points of different failure nodes of the coal and rock mass are determined .

[0023] Specifically, the first derivative analysis accurately identifies the extreme points or stationary points of the eigenvalue change, corresponding to the turning point of energy accumulation and release during the failure process of the coal and rock mass; the second derivative analysis accurately captures the inflection point of the curve, reflecting the acceleration or deceleration process of the change in the failure rate of the coal and rock mass.

[0024] Furthermore, the failure stage of the coal and rock mass is divided, including: , where is the j-th corrected demarcation time point, is the j-th demarcation time point obtained from the derivative analysis of the i-th eigenvalue fitting function, is the weight coefficient of the i-th eigenvalue; according to the corrected demarcation time point , the failure process of the coal and rock mass is divided into four stages: Stable stage I: corresponding to the crack closure stage of the coal and rock mass; Development stage II: corresponding to the linear elastic deformation stage of the coal and rock mass; Energy storage stage III: corresponding to the crack propagation stage of the coal and rock mass; Failure stage IV: corresponding to the post-peak failure stage of the coal and rock mass.

[0025] Compared with the prior art, the advantages of this application are as follows:

[0026] In improving the instability warning analysis of coal and rock mass, on the one hand, the prior art adopts simple time-domain features (such as the number of events, duration) and basic frequency-domain features (such as peak frequency, frequency center), but lacks the cumulative effect analysis and cannot reflect the failure characteristics of coal and rock mass; on the other hand, it often ignores the fluid-coal and rock mass interaction, resulting in the inability to identify the characteristic changes of different failure stages of fluid-induced coal and rock mass, and it is difficult to capture the evolution process of microcracks under the action of fluid, ultimately affecting the timeliness and accuracy of the warning.

[0027] This application improves the accuracy and timeliness of the fracture instability warning of coal and rock mass by constructing a multi-feature cumulative effect analysis system and integrating a three-stage division method of the fluid-coal and rock mass interaction mechanism.

[0028] Specifically, on the one hand, this application constructs a system containing five cumulative eigenvalues to The multi - feature system, where each feature clearly corresponds to a different aspect of the microscopic damage of coal and rock masses: the cumulative ringing feature value reflects the historical change of the number of micro - cracks; the cumulative b - feature value records the evolution process of the crack size distribution; the cumulative dominant frequency feature value characterizes the changing trend of the micro - crack propagation rate; the cumulative amplitude feature value records the cumulative effect of the energy release intensity; the acoustic emission spatial feature value describes the cumulative change in the spatial dimension; through feature accumulation, the system can "remember" the historical path of the failure of coal and rock masses, avoiding the defect in the existing technology of only focusing on instantaneous parameters and ignoring the cumulative effect, and realizing continuous and comprehensive monitoring of the progressive failure process of coal and rock masses.

[0029] On the other hand, the fluid - coal and rock mass interaction process is divided into three stages: initial adsorption, diffusion penetration, and crack propagation. Through the dual criteria of "acoustic emission event rate + frequency - band energy ratio", the system can accurately identify the whole process of fluid from surface adsorption to internal penetration and then to induced crack propagation, improving the early warning accuracy. Brief Description of the Drawings

[0030] Figure 1 is a multi - signal fusion analysis method for the fracture instability of coal and rock masses in this embodiment;

[0031] Figure 2 is a schematic diagram of the arrangement of acoustic emission sensor probes in this embodiment;

[0032] Figure 3 is the acoustic emission wavelet waveform diagram of a certain fracture event in this embodiment;

[0033] Figure 4 is the waveform diagram after wavelet transform in this embodiment;

[0034] Figure 5 is the fitting function curve in this embodiment;

[0035] Figure 6 is the first - order derivative function of the fitting function in this embodiment;

[0036] Figure 7 is the second - order derivative function of the fitting function in this embodiment;

[0037] Figure 8 is the division of the rock failure stage in this embodiment. Detailed Implementation Modes

[0038] The following describes the present application in detail with reference to the drawings in the specification and specific embodiments.

[0039] As Figure 1As shown, collect acoustic emission data during the deformation and failure process of coal and rock masses; the acoustic emission data includes acoustic emission signals and the corresponding spatial data; perform time-frequency domain processing on the acoustic emission signals to obtain the cumulative ringing characteristic value of acoustic emission , cumulative b characteristic value , cumulative main control frequency characteristic value of acoustic emission and cumulative amplitude characteristic value ; extract the cumulative energy value of acoustic emission from the spatial data as the spatial characteristic value of acoustic emission .

[0040] Fit the fitting functions of the characteristic values to respectively ; according to the fitting function , calculate the correlation coefficient matrix R between the characteristic values and ; according to the correlation coefficient matrix R, determine the weight coefficients of each characteristic value ; obtain the first derivative function and the second derivative function of the fitting function , and according to the first and the second derivative function , determine the demarcation time point during the failure process of coal and rock masses ; use the weight coefficient to correct the demarcation time point to obtain the corrected demarcation time point , and according to the corrected demarcation time point , divide the failure process of coal and rock masses into a stable period, a development period, an energy storage period, and a failure period; obtain the time range when the coal and rock masses enter the failure period as the early warning signal for the instability precursor of coal and rock masses.

[0041] Set the acoustic emission signal parameters for monitoring the deformation and failure process of coal and rock masses, establish a signal waveform and spatial parameter collection and identification model, bond acoustic emission probes to the coal and rock masses, and collect acoustic emission waveform and signal spatial parameter data; first, this embodiment is described by taking a laboratory test as an example. Prepare a 100mm×100mm×100mm CO2-water-coal sample to simulate the similar environment of a carbon storage coal pillar in underground coal mines during CO2 geological sequestration, and conduct a true triaxial loading and fracturing test on the coal sample, and monitor it by the acoustic emission method.

[0042] The parameter settings for collecting the spatial positioning characteristics of acoustic emission signals are as follows: to accurately locate the spatial position of acoustic emission signals, 2 acoustic emission probes are set on each of the loading plates in front of, below, and to the right of the true triaxial cell body, and the layout scheme of the acoustic emission probes is as Figure 2 shown.

[0043] The parameter settings for acoustic emission waveform signal collection are as follows: the working frequency range is from 125 kHz to 750 kHz; the resonant frequency is 300 kHz; the threshold value and gain are both 40 dB; the upper and lower limits of the analog filter are 1 and 400 kHz respectively, and the sampling frequency is 2 MHz;

[0044] The parameter settings for the acoustic emission waveform signal recognition model are as follows: time, channel, rise time, count, energy, duration, amplitude, average frequency, peak amplitude, back-calculated frequency, initial frequency, signal strength, absolute energy.

[0045] Next, extract the acoustic emission signal data file and its spatial parameter data characterizing the deformation and fracture events during the compression process of coal and rock masses; collect and extract the acoustic emission signal data. Each acoustic emission waveform and spatial positioning data collected by the acoustic emission monitoring system will be automatically stored as a data file in *.csv format. Each data file uses an acoustic emission waveform to record a fracture event and its spatial positioning information, and the data file of each fracture event will be processed and output as files in *.TXT and *.XLS formats for the subsequent processing of acoustic emission time-domain - frequency-domain - spatial characteristic signals.

[0046] Then, calculate and process the obtained acoustic emission data file to obtain acoustic emission time-domain signals: acoustic emission ring count eigenvalue, acoustic emission b-value eigenvalue, etc.; frequency-domain signals: acoustic emission main control frequency eigenvalue and corresponding acoustic emission amplitude eigenvalue, etc.; acoustic emission spatial eigenvalue.

[0047] Extraction of acoustic emission time-domain signals: Extract the acoustic emission ring count from the original data file obtained through experiments. Using the extracted acoustic emission ring count parameter, the acoustic emission cumulative ring count is used as the acoustic emission ring count eigenvalue.

[0048] According to the relationship between earthquake magnitude and frequency: , where M is the earthquake magnitude; N is the earthquake frequency between M + ∆M; a and b are constants. The acoustic emission b-value is calculated by replacing the magnitude M with the amplitude divided by ten, and the calculation formula is as follows: , where E is the acoustic emission absolute energy; N(E) is the number of acoustic emission hits with acoustic emission absolute energy greater than E.

[0049] In particular, there is a problem that the fixed window parameters are not suitable due to the dynamic change of the characteristics of coal and rock masses during the CO2 injection process. During the carbon storage process, the internal structure of coal and rock masses continuously changes with the processes of CO2 adsorption, diffusion, and penetration, which makes the traditional sliding window method with fixed parameters have obvious defects: too large a window will cover up key short-term changes, and too small a window will introduce too much noise. Therefore, this application realizes the automatic adjustment of window parameters according to the change of the structure of coal and rock masses by establishing a correlation function between CVR and window size, so as to adapt to the signal characteristics of different injection stages.

[0050] Specifically, set the initial and the step size , and use the sliding window algorithm to segment the acoustic emission data; calculate the cumulative variance ratio CVR of the acoustic emission signals within the current window, where CVR represents the ratio of the variance of the signal energy to the average energy within the window; calculate the optimal window of the current window according to CVR , , where γ is a positive coefficient, is the initial window length; calculate the acoustic emission b-value of each window by the least squares method according to the optimal window ; accumulate the acoustic emission b-values within each window in chronological order to obtain the cumulative b eigenvalue of acoustic emission .

[0051] , where is the variance of the energy within the calculation window; is the average value of the energy within the calculation window.

[0052] Among them, the window size is dynamically adjusted according to the ratio CVR of the signal energy variance to the average energy, so that the window length adapts to match the signal characteristics; when the signal energy fluctuates violently (large CVR value), the window length is automatically reduced to improve the time resolution and capture the information of rapidly changing microcracks; when the signal is stable (small CVR value), the window length is automatically increased to enhance the statistical stability and suppress noise interference; compared with the fixed window algorithm, the adaptive window algorithm can obtain the best signal-to-noise ratio data processing effect at different stages of the fluid (such as CO2) injection process. A non-linear mapping between the window size and the signal volatility is established through the e-exponential function relationship, avoiding the problems of over-adjustment or insufficient response.

[0053] Extraction of the acoustic emission frequency-domain signal: First, calculate the acoustic emission event rate and the proportion of band energy according to the acoustic emission data; set the first threshold and the second threshold of the acoustic emission event rate; where ; set the third threshold , the fourth threshold and the fifth threshold of the proportion of band energy in the low-frequency band, medium-frequency band and high-frequency band respectively; when the acoustic emission event rate is less than or equal to , and the proportion of band energy in the low-frequency band is greater than , it is judged as the initial adsorption stage; when the acoustic emission event rate is greater than and less than , and the proportion of band energy in the medium-frequency band is greater than , it is judged as the diffusion and penetration stage; when the acoustic emission event rate is greater than , and the proportion of the band energy in the high-frequency band is greater than When, it is determined to be the crack propagation stage.

[0054] In this embodiment, represents the critical point for the transition from surface adsorption to internal diffusion, reflecting the physical critical state where fluid molecules begin to migrate from the surface into the coal and rock mass, and the value range is 5 - 20 events / minute; characterizes the transition point from steady-state diffusion to unsteady crack propagation, and the value range is 30 - 100 events / minute; reflects the dominant position of low-frequency energy during the surface adsorption process, and the value range is 40% - 70%; represents the energy ratio of internal pore deformation and initial microcrack propagation, and the value range is 30% - 60%; characterizes the proportion of high-frequency energy in the stage of rapid crack propagation and connection, and the value range is 20% - 50%.

[0055] Then, the window lengths of the sliding windows in the initial adsorption stage, diffusion and penetration stage, and crack propagation stage are set to , and respectively; among them, ; among them, among them, The value range is 2000 - 5000 sampling points, The value range is 800 - 2000 sampling points, The value range is 200 - 800 sampling points.

[0056] According to the windows , and , the acoustic emission wave data of each stage are segmented to obtain the acoustic emission time-domain signals within multiple time windows ;

[0057] Wavelet transform is performed on the acoustic emission data within each window to obtain wavelet coefficients; the wavelet transform uses the following formula:

[0058] , where, is the wavelet coefficient, s is the scale parameter, τ is the translation parameter, and f(t) is the time-domain signal; is the conjugate function of the wavelet basis function. In particular, in this embodiment, the Daubechies wavelet family (dbN) is selected as the wavelet basis function, where N is an even number from 4 to 10. This wavelet basis function has good time-frequency localization ability and can effectively identify the non-stationary signal characteristics and mutation points of the coal and rock mass during the process of fluid (such as CO2) injection from adsorption, diffusion to crack propagation stages.

[0059] The wavelet transform has better time-frequency localization characteristics than the traditional Fourier transform and can obtain the time-domain and frequency-domain information of signals simultaneously; multi-band analysis realizes the classification and identification of coal and rock mass cracks at different scales: the low frequency mainly corresponds to large-scale structural changes, the medium frequency corresponds to mesoscopic crack propagation, and the high frequency corresponds to micro-crack initialization; by analyzing the main control frequencies and the maximum amplitudes of each frequency band, the internal structure evolution characteristics of coal and rock mass during the injection process of fluids (such as CO2) are accurately characterized.

[0060] According to the wavelet coefficients, calculate the wavelet energy distribution, and divide the wavelet energy distribution into low-frequency band, medium-frequency band and high-frequency band; in this application, the value range of the low-frequency band is 1 to 30 kHz, corresponding to the large-scale structural changes of coal and rock mass; the value range of the medium-frequency band is 30 to 100 kHz, corresponding to the mesoscopic crack propagation process; the value range of the high-frequency band is 100 to 300 kHz, corresponding to the micro-crack initialization process; this frequency band division can effectively distinguish the microscopic structural changes of coal and rock mass at different scales during the injection process of fluids (such as CO2).

[0061] Perform peak analysis on each frequency band, and extract the main control frequency of the low-frequency band and the corresponding maximum amplitude in the initial adsorption stage, the main control frequency of the medium-frequency band and the corresponding maximum amplitude in the diffusion and penetration stage, and the main control frequency of the high-frequency band and the corresponding maximum amplitude in the crack propagation stage; among them, the main control frequency represents the frequency where the peak of the wavelet energy distribution in the corresponding frequency band is located; accumulate the main control frequencies of each time window to obtain the cumulative main control frequency eigenvalue ; accumulate the maximum amplitudes of each time window to obtain the cumulative amplitude eigenvalue ; the acoustic emission wavelet waveform diagram of a certain rupture event extracted in this embodiment is as Figure 3 shown, and the waveform diagram after wavelet transform is as Figure 4 shown.

[0062] Extraction of acoustic emission spatial signals: Extract the acoustic emission energy parameters from the spatial positioning information file in the original data file, and use the acoustic emission cumulative energy as the acoustic emission spatial eigenvalue.

[0063] Use the deviation normalization method to eliminate the dimension difference for the calculated acoustic emission time-domain - frequency-domain - spatial eigenvalues, perform data normalization processing, and establish an acoustic emission time-domain - frequency-domain - spatial feature fitting function for the normalized result data;

[0064] Since the time-domain - frequency-domain - space signal responses of acoustic emission during the coal body failure process have different dimensions, in order to eliminate the dimensional differences of this information, a normalization method of deviation normalization is used to process it to obtain a dimensionless scalar.

[0065] The expression of the deviation normalization method is: ; where: is the normalized function value of the monitoring data, x is the monitoring data; k is the ordinal number of different multi-source information parameters; the value range of the monitoring data is [min, max].

[0066] Next, the acoustic emission time-domain - frequency-domain - space eigenvalue data obtained by the normalization process is fitted to construct a polynomial function such as Figure 5 The expression of the fitting function is: , where: In the function, i is 1, 2,..., 5, respectively representing the acoustic emission cumulative ringing eigenvalue ; the acoustic emission cumulative b-value eigenvalue ; the acoustic emission cumulative main control frequency eigenvalue ; the acoustic emission cumulative amplitude eigenvalue ; the acoustic emission spatial eigenvalue ; is the fitting function coefficient, is the constant term of the polynomial, representing the initial state value of the eigenvalue; j represents the highest degree of polynomial fitting.

[0067] The value of the fitting function is as follows:

[0068] In this embodiment, the fitting function parameters of the 5 acoustic emission eigenvalues are shown in the following table. R2 represents the fitting degree of the fitting function. The closer the value of R2 is to 1, the higher the fitting degree is proved.

[0069] As in the following table the fitting degree indexes are 0.9869, 0.9954, 0.9415, 0.9749, 0.9108 respectively, indicating that the fitting degree of the constructed polynomial function is relatively ideal. See Table 1 for details;

[0070] Table 1 Fitting function parameter table

[0071]

[0072] Analyze the characteristics of the closely related quantities of the acoustic emission time-domain - frequency-domain - space eigenvalue fitting function to obtain the influence weight of the acoustic emission time-domain - frequency-domain - space eigenvalue during the failure process of coal and rock mass;

[0073] First, based on the minimum domain of the monitoring information in the above polynomial fitting function, determine the domain of the sample data matrix, and then obtain the acoustic emission time-frequency-space eigenvalue sample matrix F:

[0074]

[0075] The correlation coefficient matrix R is calculated from the sample matrix F through the following calculation formula:

[0076] The calculation formula is:

[0077] ; where: is the original variable and The correlation coefficient of, and are respectively and The average value of i, j = 1, 2,..., n; and , the correlation coefficient matrix R is:

[0078] ;

[0079] Calculate the eigenvalues to of the correlation coefficient matrix R and the cumulative contribution rate, and then calculate the contribution rate and cumulative contribution rate of the main components. ; ; where: E is the unit matrix; m is the number of eigenvalues that determine the information of the main components, that is, the number of main components, as shown in Table 2.

[0080] Table 2 Matrix eigenvalues and cumulative contribution rates

[0081]

[0082] To ensure that the main components can reflect the information of the original variables to the greatest extent, it is required that the cumulative contribution rate reaches more than 80%. As can be seen from the above table, the cumulative contribution rate of the eigenvalue λ is 84.136%.

[0083] The eigenvalues of the correlation coefficient matrix R and the eigenvectors corresponding to the eigenvalues are combined to form the factor loading matrix ρ, that is, the factor loading matrix ρ is determined by the eigenvalues.

[0084] Therefore, the eigenvalue is 5.885, and the factor loading matrix ρ is [0.986, 0.839, 0.688, 0.991, 0.945, 0.958, 0.985].

[0085] According to the relationship between the eigenvalue and the factor loading matrix ρ, the function coefficient matrix is obtained. , where : ; From the coefficient matrix , the principal component analysis function F(x) is obtained:

[0086] ; F(x) can be regarded as 's comprehensive variable, which mainly explains the comprehensive effect generated by five acoustic emission characteristic parameters: acoustic emission ring count, acoustic emission b value, acoustic emission main control frequency and amplitude, and acoustic emission spatial eigenvalue.

[0087] The coefficients of F(x) are the influence weights of the characteristic value parameters in the time domain - frequency domain - space of acoustic emission. Therefore, the order of the variables reflecting the specimen failure from strong to weak is as follows: ;

[0088] Solve the derivative function of the acoustic emission time domain - frequency domain - space characteristic fitting function, and judge the demarcation time point of the coal and rock mass in different failure stages according to the concavity, convexity and change of increase and decrease of the derivative function curve;

[0089] Solve the derivative function of the acoustic emission time domain - frequency domain - space characteristic value fitting function, and the first-order derivative function of the fitting function is obtained as Figure 6 and Two The second-order derivative function is as Figure 7 . When the derivative function curve intersects the x-axis, the value of the derivative function is 0, and this intersection point is the stationary point or inflection point. Analyze and judge the demarcation time point of the coal body in different failure stages according to the concavity, convexity and change of increase and decrease characteristics of the derivative function curve;

[0090] Perform weighted correction on the demarcation time point of different failure stages through the obtained extreme points, inflection points of the derivative function and the results of principal component analysis: , where is the j-th corrected demarcation time point, is the j-th demarcation time point obtained from the derivative analysis of the i-th eigenvalue fitting function, is the weight coefficient of the i-th eigenvalue; According to the corrected demarcation time point , the failure process of the coal and rock mass is divided into four stages: stable period (I): corresponds to the crack closure stage of the coal and rock mass; development period (II): corresponds to the linear elastic deformation stage of the coal and rock mass; energy storage period (III): corresponds to the crack propagation stage of the coal and rock mass; failure period (IV): corresponds to the post-peak failure stage of the coal and rock mass;

[0091] The boundary time points of different failure stages are weighted and corrected by the method of averaging through the factor loading matrix based on the obtained extreme points, inflection points of the derivative function and the corresponding principal component analysis results, as shown in the following table. Furthermore, the stage division of the precursor characteristics during the rock failure process is carried out for the stable period I to the failure period IV, as shown in Table III, and the coal body failure stage diagram is obtained as Figure 8 shown;

[0092] Table III Rock failure stage division table

[0093]

[0094] Based on the stage division results, the time range when the coal and rock mass enters the failure period IV is used as the key discriminant information for the early warning of rock instability precursors, as shown in Table IV.

[0095] Table IV Relationship table between precursor limit and bearing limit of failure stage

[0096]

[0097] For example Figure 8 shown, the division of each stage I - IV is combined with the stress - strain curve of the coal body. As can be seen from Table IV, during the failure process of the coal body, the time occupied by stage II and stage III is relatively large. The coal body transitions from the elastic stage to the crack propagation stage, which is the energy storage stage during the coal body failure process. While the time proportions of stage I and stage IV are relatively small. Stage IV represents the process of the development and expansion of internal cracks in the coal body and gradually reaching the bearing limit and rapid failure, corresponding to the crack propagation and post - peak failure stages of the coal body failure. Therefore, it is necessary to pay attention to strengthening the monitoring and early warning of stage III and stage IV.

[0098] The present application and its implementation manners are schematically described above. The description is not restrictive. Without departing from the spirit or basic characteristics of the present application, the present application can be implemented in other specific forms. What is shown in the drawings is only one of the implementation manners of the present application, and the actual structure is not limited thereto. Therefore, if those of ordinary skill in the art are inspired by it and design similar structural manners and embodiments without creative efforts without departing from the purpose of this creation, they shall fall within the protection scope of the present application. In addition, the term "including" does not exclude other elements or steps, and the term "a" before an element does not exclude including "multiple" such elements. The terms first, second, etc. are used to represent names and do not represent any specific order.

Claims

1. A multi-signal fusion analysis method for the fracture instability of coal and rock masses, characterized in that, Including: Collecting acoustic emission data during the deformation and failure process of coal and rock masses; the acoustic emission data includes acoustic emission signals and spatial data corresponding to the acoustic emission signals; Perform time-frequency domain processing on the acoustic emission signal to obtain the cumulative ringing eigenvalue of the acoustic emission , the cumulative b eigenvalue , the cumulative dominant frequency eigenvalue of the acoustic emission and the cumulative amplitude eigenvalue ; Extract the cumulative energy value of the acoustic emission from the spatial data as the spatial eigenvalue of the acoustic emission ; Fit the eigenvalues separately to the eigenvalue of the fitting function ; According to the fitting function , calculate the correlation coefficient matrix R between the eigenvalues and ; determine the weight coefficients of each eigenvalue according to the correlation coefficient matrix R; Obtain the fitting function of the first derivative function and the second derivative function and, based on the first and second derivative functions , determine the demarcation time point during the failure process of coal and rock mass ; Using the weight coefficient to correct the demarcation time point and obtain the corrected demarcation time point . According to the corrected demarcation time point , the failure process of coal and rock mass is divided into a stable period, a development period, an energy storage period and a failure period; Obtaining the time range when the coal and rock masses enter the failure period as the early warning signal for coal and rock mass instability precursors.

2. The method for multi-signal fusion analysis of coal and rock mass fracture instability according to claim 1, characterized in that: Cumulative ringing eigenvalue It represents the cumulative value of the ringing count during the acoustic emission process and reflects the number of microcracks inside the coal and rock mass; Cumulative b eigenvalue It represents the cumulative value of the slope coefficient of the amplitude-frequency distribution of acoustic emission signals, reflecting the change in the distribution of the internal crack scale of coal and rock masses during the interaction between fluid and coal and rock masses; Cumulative main control frequency eigenvalue It represents the time-sequential cumulative value of the frequency corresponding to the maximum amplitude in the acoustic emission signal spectrum, reflecting the propagation rate of microcracks inside the coal and rock mass; Cumulative amplitude eigenvalue It represents the time-sequential cumulative value of the maximum amplitude value of the acoustic emission signal, reflecting the intensity change of the energy released by the microcracks inside the coal and rock mass.

3. The method for multi-signal fusion analysis of coal and rock mass fracture instability according to claim 2, characterized in that: Obtain the cumulative b eigenvalue , including: Extract the acoustic emission absolute energy E from the collected acoustic emission data, as well as the number of hits with absolute energy greater than E ; Establish the relationship between the acoustic emission absolute energy E and the hit count as follows: , where a and b are constants; Set the initial and step size , and use the sliding window algorithm to segment the acoustic emission data; Calculating the cumulative variance ratio CVR of the acoustic emission signals within the current window, where CVR represents the ratio of the signal energy variance to the average energy within the window; Calculate the optimal window of the current window according to the CVR , , where γ is a positive coefficient; According to the optimal window , calculate the acoustic emission b value of each window by the least squares method; Accumulate the acoustic emission b values in each window in chronological order to obtain the cumulative acoustic emission b eigenvalue .

4. The method for multi-signal fusion analysis of coal and rock mass fracture instability according to claim 3, characterized in that: Cumulative main control frequency eigenvalue and cumulative amplitude eigenvalue , including: According to the acoustic emission data, dividing the process of interaction between fluid and coal and rock masses into an initial adsorption stage, a diffusion and penetration stage, and a crack propagation stage; Set the window lengths of the sliding windows for the initial adsorption stage, diffusion penetration stage, and fracture propagation stage to be , and ; where ; According to the window , and , the acoustic emission wave data in each stage is segmented to obtain the acoustic emission time domain signals within multiple time windows ; Performing wavelet transform on the acoustic emission data within each window to obtain wavelet coefficients; Calculating the wavelet energy distribution based on the wavelet coefficients and dividing the wavelet energy distribution into a low-frequency band, a medium-frequency band, and a high-frequency band; Perform peak analysis on each frequency band to extract the dominant frequency of the low-frequency band in the initial adsorption stage and the corresponding maximum amplitude , the dominant frequency of the middle-frequency band in the diffusion and penetration stage and the corresponding maximum amplitude , the dominant frequency of the high-frequency band in the crack propagation stage and the corresponding maximum amplitude ; among them, the dominant frequency represents the frequency at which the peak of the wavelet energy distribution in the corresponding frequency band is located; Accumulate the master frequencies of each time window , and obtain the cumulative master frequency eigenvalue ; Accumulate the maximum amplitudes of each time window to obtain the cumulative amplitude eigenvalue .

5. The method for multi-signal fusion analysis of coal and rock mass fracture instability according to claim 4, characterized in that: Dividing the process of interaction between fluid and coal and rock masses into an initial adsorption stage, a diffusion and penetration stage, and a crack propagation stage, including: Calculating the acoustic emission event rate and the proportion of band energy according to the acoustic emission data; Set a first threshold for the acoustic emission event rate and a second threshold ; wherein ; Set the third threshold for the proportion of the band energy of the low-frequency band, the medium-frequency band, and the high-frequency band respectively , the fourth threshold and the fifth threshold ; When the acoustic emission event rate is less than or equal to , and the proportion of the band energy in the low-frequency band is greater than ε3, it is judged as the initial adsorption stage; When the acoustic emission event rate is greater than and less than , and the proportion of the band energy in the intermediate frequency band is greater than , it is judged as the diffusion penetration stage; When the acoustic emission event rate is greater than , and the proportion of the frequency band energy in the high-frequency band is greater than , it is judged as the crack propagation stage.

6. The method for multi-signal fusion analysis of coal and rock mass fracture instability according to any one of claims 2 to 5, characterized in that: Fit the eigenvalues separately to eigenvalue of the fitting function , including: ; where correspond to five eigenvalues respectively, x is the time variable, representing the duration from the start of the interaction between the fluid and the coal-rock mass to the current time, is the j-th coefficient of the i-th eigenvalue fitting function; is the constant term of the polynomial, representing the initial state value of the eigenvalue; j represents the highest degree of polynomial fitting.

7. The method for multi-signal fusion analysis of coal and rock mass fracture instability according to claim 6, characterized in that: Determining the weight coefficients of each eigenvalue, including: Set the fitting function The common valid time interval is used as the minimum domain; According to the minimum domain, calculate the eigenvalues at the sampling time points by fitting the function as sample data ; Constructing an acoustic emission eigenvalue sample matrix F according to the sample data; Calculate the eigenvalues based on the acoustic emission eigenvalue sample matrix F to calculate the correlation coefficient matrix R, and the element in matrix R represents the correlation coefficient between the eigenvalue and ; Calculate the matrix eigenvalues of the correlation coefficient matrix R to , and the corresponding eigenvectors ; Calculate the contribution rate and cumulative contribution rate of the eigenvalues of each matrix, and select the eigenvalues of the matrix whose cumulative contribution rate is greater than the threshold ; ; According to the selected matrix eigenvalues and the corresponding eigenvectors , a factor loading matrix ρ is constructed, where ; Determine the eigenvalue according to the factor loading matrix ρ to weight coefficient , .

8. The method for multi-signal fusion analysis of coal and rock mass fracture instability according to claim 7, characterized in that: The expression is as follows: ; wherein, is the eigenvalue and is the correlation coefficient of ; and , and are respectively and the average values of represents the value of the eigenvalue at the k-th time point, represents the value of the eigenvalue at the k-th time point; k represents the serial number of the time point, .

9. The method for multi-signal fusion analysis of coal and rock mass fracture instability according to claim 8, characterized in that: Determining the demarcation time point during the failure process of coal and rock masses, including: Calculate the first derivative function at the time point, which is used as the extreme point or stationary point of the fitting function; Calculate the second derivative function at the time point as the inflection point of the fitting function; According to the first derivative function and the second derivative function , as well as the extreme points or stationary points, and inflection points, determine the boundary time points of different failure nodes of coal and rock masses .

10. The method for multi-signal fusion analysis of coal and rock mass fracture instability according to claim 9, characterized in that: Correcting the boundary time point using a weight coefficient, including: ; wherein, is the j-th corrected boundary time point, is the j-th boundary time point obtained by analyzing the derivative of the i-th eigenvalue fitting function, is the weight coefficient of the i-th eigenvalue.

Images