Coal rock mass fracture precursor signal identification method and device, medium and equipment

Abstract

The invention discloses a coal and rock mass fracture precursor signal identification method and device, a medium and equipment, and the method comprises the steps: carrying out the noise reduction of an original voltage signal corresponding to an underground sound signal and / or a vibration signal, and obtaining a denoised signal; performing time-frequency analysis on the denoised signal to obtain a time-frequency spectrum, converting the time-frequency spectrum into an energy spectrum, and constructing a three-dimensional spectrum based on the time-frequency spectrum and the energy spectrum; calculating a frequency barycentric coordinate, a three-dimensional energy entropy and a frequency band energy ratio in the three-dimensional energy based on the three-dimensional map; a frequency barycentric coordinate, a three-dimensional energy entropy and a frequency band energy ratio in the three-dimensional energy are respectively used for constructing a characteristic time sequence according to a time sequence, smoothing processing is carried out, and a frequency barycentric coordinate characteristic sequence, an energy entropy characteristic sequence and a frequency band energy ratio characteristic sequence after smoothing processing are analyzed with respective corresponding preset precursor characteristic sequences. And obtaining an early warning level according to an analysis result. According to the invention, the accuracy and reliability of precursor signal identification and early warning are improved.

Description

Technical Field

[0001] This invention relates to the field of geotechnical engineering technology, and in particular to a method, apparatus, medium and equipment for identifying precursor signals of coal and rock mass fracturing. Background Technology

[0002] Instability and fracturing of coal and rock masses (such as rockbursts, coal and gas outbursts, and roof collapses) are major hazards in coal mine safety production. Extensive theoretical and practical research shows that before instability, coal and rock masses undergo the initiation, propagation, and connection of internal microcracks, accompanied by a large number of acoustic emission (AE) or microseismic (MS) signals. These signals contain rich information about precursory ruptures; therefore, analyzing these signals to identify precursory ruptures is crucial for accurate disaster early warning.

[0003] Existing methods for identifying precursor signals of coal and rock fracturing mainly rely on single-parameter or simple joint analysis of signal time-domain parameters (such as ring count, energy, and duration) or frequency-domain parameters (such as dominant frequency and spectral centroid). However, these methods have the following significant problems and drawbacks: 1. Low Accuracy of Precursor Feature Extraction: Single-dimensional analysis in the time or frequency domain cannot fully characterize the non-stationary and nonlinear acoustic emission / microseismic signal features. Time-domain analysis loses frequency structure information, while frequency-domain analysis (such as FFT) loses time evolution information, making it difficult to capture the essential transient and abrupt characteristics of precursor signals, resulting in low extraction accuracy. Because rupture precursor signals are often weak in energy, have complex frequency components, and frequently overlap with background noise frequency bands, simple time-frequency analysis cannot construct high-resolution, high-definition feature maps, leading to insufficient sensitivity to minute changes in precursors and low extraction accuracy.

[0004] 2. Weak anti-interference ability: The downhole environment is complex with noise (such as mechanical vibration and electromagnetic interference). Traditional methods are difficult to effectively distinguish between effective signals and noise generated by actual rock mass fractures, which can easily lead to false alarms or missed alarms.

[0005] Therefore, there is an urgent need for a high-precision precursor signal identification method with strong noise resistance. Summary of the Invention

[0006] In view of this, the present invention provides a method for identifying precursor signals of coal and rock mass fracturing, the main purpose of which is to solve the problems of low identification accuracy and weak anti-interference ability of existing methods for identifying precursor signals of coal and rock mass fracturing.

[0007] According to one aspect of this application, a method for identifying precursor signals of coal and rock mass fracturing is provided, the method comprising: Obtain the original voltage signal corresponding to the sound signal and / or vibration signal downhole, and perform noise reduction processing on the original voltage signal to obtain the noise-reduced signal; The denoised signal is subjected to wavelet transform to obtain the wavelet transform time spectrum, the wavelet transform time spectrum is converted into an energy spectrum, and a three-dimensional map is constructed based on the wavelet transform time spectrum and the energy spectrum. Based on the three-dimensional spectrum, calculate the frequency centroid coordinates in the three-dimensional energy, and calculate the three-dimensional energy entropy and frequency band energy ratio based on the three-dimensional spectrum; The frequency centroid coordinates, the three-dimensional energy entropy, and the frequency band energy ratio in the three-dimensional energy are respectively constructed into feature time series in time order and smoothed to obtain smoothed frequency centroid coordinate feature series, energy entropy feature series, and frequency band energy ratio feature series. The smoothed frequency centroid coordinate feature sequence, energy entropy feature sequence, and frequency band energy ratio feature sequence are analyzed with their respective preset precursor feature sequences. Based on the analysis results, the warning level is obtained.

[0008] Optionally, the step of performing noise reduction processing on the original voltage signal to obtain the noise-reduced signal includes: The original voltage signal is decomposed using an adaptive noise set empirical mode decomposition method to obtain multiple intrinsic mode function components and residual terms. The effective signals of the intrinsic mode function components are then filtered to obtain effective component signals and invalid component signals. The effective component signal and the invalid component signal are quantized using a semi-soft thresholding function respectively; The residual term, the quantized effective component signal, and the invalid component signal are reconstructed to obtain the denoised signal.

[0009] Optionally, the step of filtering the intrinsic mode function components to obtain valid component signals and invalid component signals includes: Calculate the correlation coefficient and variance contribution rate between each intrinsic mode function component and the original voltage signal; The intrinsic mode function components with correlation coefficients greater than a preset coefficient threshold and variance contribution rates greater than a preset contribution rate are considered as valid component signals, while the remaining intrinsic mode function components are considered as invalid component signals.

[0010] Optionally, performing wavelet transform on the denoised signal to obtain the wavelet transform time-frequency spectrum, and converting the wavelet transform time-frequency spectrum into an energy spectrum, includes: Perform continuous wavelet transform on the denoised signal to obtain wavelet coefficients, and calculate the instantaneous frequency of the wavelet coefficients; The wavelet coefficients and their instantaneous frequencies are calculated, and the wavelet coefficients are mapped from the scale-time plane to the frequency-time plane to obtain the wavelet transform spectrum. The square of the absolute value of the complex time spectrum in the wavelet transform time spectrum is taken as the energy spectrum corresponding to the wavelet transform time spectrum.

[0011] Optionally, the frequency centroid coordinates in the three-dimensional energy are calculated using the following method:

[0012] in, P ( t m , f k ) for time t m and frequency f k Energy intensity at that location f c Let M be the frequency centroid coordinate in the three-dimensional energy, M be the total number of time discrete points, and K be the total number of frequency discrete points.

[0013] Optionally, calculating the three-dimensional energy entropy and band energy ratio based on the three-dimensional spectrum includes: Based on the three-dimensional map, the probability of the three-dimensional energy distribution is calculated to obtain the probability of the energy intensity distribution at the time and frequency points; The three-dimensional energy entropy is obtained by calculating the probability of the energy intensity distribution at time and frequency points. Based on a preset frequency segmentation threshold, the time spectrum is divided into a first frequency band and a second frequency band. Based on the energy intensity in the three-dimensional spectrum, the total low-frequency energy of the first frequency band and the total high-frequency energy of the second frequency band are calculated respectively. The ratio of the total high-frequency energy to the total low-frequency energy is taken as the band energy ratio.

[0014] Optionally, the step of analyzing the smoothed frequency centroid coordinate feature sequence, energy entropy feature sequence, and frequency band energy ratio feature sequence with their respective preset precursor feature sequences, and obtaining the warning level based on the analysis results, includes: Calculate the first distance between the smoothed frequency centroid coordinate feature sequence and its corresponding preset precursor feature sequence, determine the trend of the frequency centroid coordinate within a preset time period based on the smoothed frequency centroid coordinate feature sequence, and determine the warning value of the frequency centroid coordinate based on the trend and the first distance. Calculate the second distance between the smoothed energy entropy feature sequence and its corresponding preset precursor feature sequence, determine the energy entropy decrease mode based on the smoothed energy entropy feature sequence, and determine the energy entropy warning value based on the decrease mode and the second distance; Calculate the third distance between the smoothed frequency band energy ratio feature sequence and its corresponding preset precursor feature sequence, determine the inflection point mode of the frequency band energy ratio based on the smoothed frequency band energy ratio feature sequence, and determine the warning value of the frequency band energy ratio according to the inflection point mode and the third distance. The total warning level is calculated based on the warning values ​​of the frequency centroid coordinates, the energy entropy, and the frequency band energy ratio, as well as their respective weights.

[0015] According to another aspect of this application, a device for identifying precursor signals of coal and rock mass fracturing is provided, comprising: The noise reduction module is used to acquire the original voltage signal corresponding to the sound signal and / or vibration signal downhole, and to perform noise reduction processing on the original voltage signal to obtain the noise-reduced signal. A three-dimensional atlas construction module is used to perform wavelet transform on the denoised signal to obtain the wavelet transform time spectrum, convert the wavelet transform time spectrum into an energy spectrum, and construct a three-dimensional atlas based on the wavelet transform time spectrum and the energy spectrum; The eigenvalue calculation module is used to calculate the frequency centroid coordinates in the three-dimensional energy based on the three-dimensional spectrum, and to calculate the three-dimensional energy entropy and frequency band energy ratio based on the three-dimensional spectrum. The feature sequence acquisition module is used to construct feature time series of the frequency centroid coordinates, the three-dimensional energy entropy and the frequency band energy ratio in the three-dimensional energy according to the time sequence and perform smoothing processing to obtain smoothed frequency centroid coordinate feature series, energy entropy feature series and frequency band energy ratio feature series. The analysis module is used to analyze the smoothed frequency centroid coordinate feature sequence, energy entropy feature sequence, and frequency band energy ratio feature sequence with their respective preset precursor feature sequences, and obtain the warning level based on the analysis results.

[0016] Optionally, the noise reduction module is further configured to: The original voltage signal is decomposed using an adaptive noise set empirical mode decomposition method to obtain multiple intrinsic mode function components and residual terms. The effective signals of the intrinsic mode function components are then filtered to obtain effective component signals and invalid component signals. The effective component signal and the invalid component signal are quantized using a semi-soft thresholding function respectively; The residual term, the quantized effective component signal, and the invalid component signal are reconstructed to obtain the denoised signal.

[0017] Optionally, the noise reduction module is further configured to: Calculate the correlation coefficient and variance contribution rate between each intrinsic mode function component and the original voltage signal; The intrinsic mode function components with correlation coefficients greater than a preset coefficient threshold and variance contribution rates greater than a preset contribution rate are considered as valid component signals, while the remaining intrinsic mode function components are considered as invalid component signals.

[0018] Optionally, the three-dimensional map construction module is further used for: Perform continuous wavelet transform on the denoised signal to obtain wavelet coefficients, and calculate the instantaneous frequency of the wavelet coefficients; The wavelet coefficients and their instantaneous frequencies are calculated, and the wavelet coefficients are mapped from the scale-time plane to the frequency-time plane to obtain the wavelet transform spectrum. The square of the absolute value of the complex time spectrum in the wavelet transform time spectrum is taken as the energy spectrum corresponding to the wavelet transform time spectrum.

[0019] Optionally, the frequency centroid coordinates in the three-dimensional energy are calculated using the following method:

[0020] in, P ( t m , f k ) for time t m and frequency f k Energy intensity at that location f c Let M be the frequency centroid coordinate in the three-dimensional energy, M be the total number of time discrete points, and K be the total number of frequency discrete points.

[0021] Optionally, the eigenvalue calculation module is further configured to: Based on the three-dimensional map, the probability of the three-dimensional energy distribution is calculated to obtain the probability of the energy intensity distribution at the time and frequency points; The three-dimensional energy entropy is obtained by calculating the probability of the energy intensity distribution at time and frequency points. Based on a preset frequency segmentation threshold, the time spectrum is divided into a first frequency band and a second frequency band. Based on the energy intensity in the three-dimensional spectrum, the total low-frequency energy of the first frequency band and the total high-frequency energy of the second frequency band are calculated respectively. The ratio of the total high-frequency energy to the total low-frequency energy is taken as the frequency band energy ratio.

[0022] Optionally, the analysis module is further configured to: Calculate the first distance between the smoothed frequency centroid coordinate feature sequence and its corresponding preset precursor feature sequence, determine the trend of the frequency centroid coordinate within a preset time period based on the smoothed frequency centroid coordinate feature sequence, and determine the warning value of the frequency centroid coordinate based on the trend and the first distance. Calculate the second distance between the smoothed energy entropy feature sequence and its corresponding preset precursor feature sequence, determine the energy entropy decrease mode based on the smoothed energy entropy feature sequence, and determine the energy entropy warning value based on the decrease mode and the second distance; Calculate the third distance between the smoothed frequency band energy ratio feature sequence and its corresponding preset precursor feature sequence, determine the inflection point mode of the frequency band energy ratio based on the smoothed frequency band energy ratio feature sequence, and determine the warning value of the frequency band energy ratio according to the inflection point mode and the third distance. The total warning level is calculated based on the warning values ​​of the frequency centroid coordinates, the energy entropy, and the frequency band energy ratio, as well as their respective weights.

[0023] According to another aspect of this application, a storage medium is provided, wherein at least one executable instruction is stored therein, the executable instruction causing a processor to perform the operation corresponding to the above-described method for identifying precursor signals of coal and rock mass fracturing.

[0024] According to another aspect of this application, a computer device is provided, comprising: a processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, which causes the processor to perform the operation corresponding to the above-described method for identifying precursor signals of coal and rock mass fracturing.

[0025] By employing the above-described technical solutions, the technical solutions provided by the embodiments of the present invention have at least the following advantages: This application provides a method, device, medium, and equipment for identifying precursor signals of coal and rock mass fracturing. By denoising the original voltage signal, performing time-frequency analysis and energy conversion on the denoised signal, a high-resolution time-frequency-energy three-dimensional distribution map is constructed. Based on the three-dimensional distribution map, multi-dimensional feature indicators that sensitively reflect the evolution of coal and rock mass fracturing are extracted. The multi-dimensional feature indicators and their corresponding preset precursor features are analyzed. Based on the analysis results, the warning level is obtained. Through time-frequency analysis and time-frequency-energy three-dimensional joint feature extraction, weak, early precursor features that traditional methods cannot detect can be captured, significantly improving identification accuracy and potentially obtaining a longer warning lead time. The denoising process can effectively distinguish real fracturing signals from various underground environmental noises, greatly reducing false alarm and missed alarm rates, and improving the accuracy and reliability of the warning.

[0026] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0027] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart of a method for identifying precursor signals of coal and rock mass fracturing according to an embodiment of this application is shown; Figure 2 The original voltage signal diagram of a method for identifying precursor signals of coal and rock mass fracturing according to an embodiment of this application is shown. Figure 3 Another flowchart of a method for identifying precursor signals of coal and rock mass fracturing provided in an embodiment of this application is shown; Figure 4 The diagram shows a characteristic sequence of a method for identifying precursor signals of coal and rock mass fracturing according to an embodiment of this application. Figure 5 This paper shows a structural block diagram of a coal and rock mass fracturing precursor signal identification device provided in an embodiment of this application; Figure 6A schematic diagram of the structure of a computer device provided in an embodiment of the present invention is shown.

[0028] in, Figure 5 In Chinese: 502 - Noise Reduction Module; 504 - 3D Map Construction Module; 506 - Eigenvalue Calculation Module; 508 - Feature Sequence Acquisition Module; 510 - Analysis Module; Figure 6 In Chinese: 602 - Processor; 604 - Communication interface; 606 - Memory; 608 - Communication bus; 610 - Program. Detailed Implementation

[0029] The present invention will be described in detail below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in the embodiments of the present invention can be combined with each other.

[0030] To further illustrate the technical means and effects adopted by the present invention to achieve the intended purpose, the specific embodiments, structures, features, and effects according to the present invention will be described in detail below with reference to the accompanying drawings and preferred embodiments. In the following description, different "an embodiment" or "an embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0031] To address the problems of low identification accuracy and weak anti-interference ability of existing methods for identifying precursor signals of coal and rock mass fracturing, this application provides a method for identifying precursor signals of coal and rock mass fracturing, such as... Figure 1 As shown, the method includes: 102: Obtain the original voltage signal corresponding to the sound signal and / or vibration signal downhole, perform noise reduction processing on the original voltage signal, and obtain the noise-reduced signal; 104: Perform wavelet transform on the denoised signal to obtain the wavelet transform time spectrum, convert the wavelet transform time spectrum into an energy spectrum, and construct a three-dimensional spectrum based on the wavelet transform time spectrum and energy spectrum; 106: Based on the three-dimensional spectrum, calculate the frequency centroid coordinates in the three-dimensional energy, and calculate the three-dimensional energy entropy and frequency band energy ratio based on the three-dimensional spectrum; 108: Construct feature time series of frequency centroid coordinates, three-dimensional energy entropy, and frequency band energy ratio in three-dimensional energy according to time order and perform smoothing to obtain smoothed frequency centroid coordinate feature series, energy entropy feature series, and frequency band energy ratio feature series; 110: Analyze the smoothed frequency centroid coordinate feature sequence, energy entropy feature sequence, and frequency band energy ratio feature sequence with their respective preset precursor feature sequences, and obtain the warning level based on the analysis results.

[0032] Specifically, the raw voltage signal is acquired through an array of acoustic emission / microseismic sensors deployed downhole. x ( t The original voltage signal is denoised. For example, a joint denoising method combining fully adaptive noise set empirical mode decomposition (ICEEMDAN) and wavelet semi-soft threshold function is used to preprocess x(t) to obtain the denoised signal with good denoising effect.

[0033] The denoised signal is subjected to sparse wavelet transform to obtain the wavelet transform time spectrum. The wavelet transform time spectrum is then converted into an energy spectrum. Based on the time spectrum and energy spectrum, a three-dimensional map is constructed. Feature factors are extracted based on the three-dimensional map. By analyzing the feature factors, weak and early precursor features that cannot be detected by traditional methods can be captured, which significantly improves the identification accuracy and is expected to obtain a longer warning lead time.

[0034] The extracted feature factors include feature indices (frequency and barycentric coordinates in three-dimensional energy). fc 3D energy entropy H 3D Bandwidth energy ratio R HL (etc.), construct feature time series of feature factors in chronological order and perform smoothing, for example, using a Savitzky-Golay filter for smoothing:

[0035] F ( t ) is the original feature sequence (e.g. fc ( t The feature sequence after being smoothed by the Savitzky-Golay filter. S G (.) represents the Savitzky-Golay filtering operation, which achieves smoothing by performing polynomial least squares fitting within a local window.

[0036] This method analyzes the smoothed frequency centroid coordinate feature sequence, energy entropy feature sequence, and frequency band energy ratio feature sequence with their respective preset precursor feature sequences. For example, it calculates the similarity between these sequences and their preset precursor feature sequences to determine their patterns. Based on similarity and patterns, it determines the warning value for each feature factor and then uses a weighted voting fusion decision model to determine the overall warning level. By combining time-frequency analysis and time-frequency-energy three-dimensional joint feature extraction, it achieves accurate and robust identification of weak, early rupture precursor signals, providing more reliable technical support for coal mine disaster early warning. This method is not only suitable for acoustic emission / microseismic monitoring in coal mines but can also be widely applied to rock mass stability monitoring and disaster early warning in metal mines, hydropower projects, and tunnel geotechnical engineering, showing promising prospects for widespread application.

[0037] This application provides a method for identifying precursor signals of coal and rock mass fracturing. Compared with existing technologies, this method denoises the original voltage signal, performs time-frequency analysis and energy conversion on the denoised signal, and constructs a high-resolution time-frequency-energy three-dimensional distribution map. Based on the three-dimensional distribution map, it extracts multi-dimensional feature indicators that can sensitively reflect the evolution process of coal and rock mass fracturing. It analyzes the multi-dimensional feature indicators and their corresponding preset precursor features, and obtains the warning level based on the analysis results. Through time-frequency analysis and time-frequency-energy three-dimensional joint feature extraction, it can capture weak, early precursor features that traditional methods cannot detect, significantly improving the identification accuracy and potentially obtaining a longer warning lead time. The denoising process can effectively distinguish real fracturing signals from various underground environmental noises, greatly reducing the false alarm rate and missed alarm rate, and improving the accuracy and reliability of the warning.

[0038] In one embodiment, the original voltage signal is denoised to obtain a denoised signal, including: The original voltage signal is decomposed using an adaptive noise ensemble empirical mode decomposition method to obtain multiple intrinsic mode function components and residual terms. The effective signals of the intrinsic mode function components are then filtered to obtain effective component signals and invalid component signals. A semi-soft thresholding function is used to quantize the effective component signal and the invalid component signal respectively; The residual term, the quantized effective component signal, and the invalid component signal are reconstructed to obtain the denoised signal.

[0039] Specifically, the raw voltage signal is acquired through an array of acoustic emission / microseismic sensors deployed downhole. x ( t ),like Figure 2As shown, a joint denoising method based on an improved fully adaptive noise ensemble empirical mode decomposition (ICEEMDAN) combined with a wavelet semi-soft thresholding function is used to denoise the noise. x ( t Preprocessing is performed.

[0040] 1. Using ICEEMDAN x ( t It is decomposed into a series of intrinsic mode function components (IMFs). c i ( t ) and a residual r n ( t ):

[0041] x ( t () represents the original time-domain signal acquired. c i ( t ) is the first decomposition obtained i Each intrinsic mode function component r n ( t The residual term after decomposition represents the trend or mean of the signal. n This represents the total number of intrinsic mode function components.

[0042] 2. Calculate the correlation coefficient between each intrinsic mode function component and the original voltage signal. ρ i and variance contribution rate η i The intrinsic mode function (IMF) components with correlation coefficients greater than a preset threshold and variance contribution rates greater than a preset contribution rate are considered valid IMF components, while the remaining IMF components are considered invalid IMF components. For example, based on a set threshold (such as...),... ρ i >0.1 and η i >1%), filtering out the IMF component set that mainly contains valid signals { C valid The remaining IMF components are treated as invalid component signals.

[0043] 3. For each component in the effective signal, a continuously differentiable semi-soft thresholding function is applied for quantization to further suppress residual noise:

[0044] For the thresholded first iThere are 1 effective component signal, where sign() is the sign function and || is the absolute value function. λ The threshold parameter is estimated based on the noise level. μ As an adjustment factor, it controls the smoothness of the transition from the soft threshold to the hard threshold in the function, and is usually taken as 2 to 5. ) + This is for operations that take positive values; that is, if the result inside the parentheses is negative, then it is taken as 0.

[0045] The invalid component signal is quantized using a continuously differentiable semi-soft threshold function.

[0046] 4. Reconstruct the denoised signal by combining the processed effective component signal, invalid component signal, and the unprocessed residual. s ( t ):

[0047] s ( t The signal is the clean time-domain signal after joint noise reduction processing. c j ( t The signal that is determined to be an invalid component dominated by noise is retained during reconstruction to avoid distortion of the valid signal.

[0048] In one embodiment, wavelet transform is performed on the denoised signal to obtain the wavelet transform time spectrum, and the wavelet transform time spectrum is converted into an energy spectrum, including: Perform continuous wavelet transform on the denoised signal to obtain wavelet coefficients, and calculate the instantaneous frequency of the wavelet coefficients; The wavelet transform spectrum is obtained by calculating the wavelet coefficients and their instantaneous frequencies, mapping them from the scale-time plane to the frequency-time plane. The square of the absolute value of the complex time spectrum in the wavelet transform time spectrum is taken as the energy spectrum corresponding to the wavelet transform time spectrum.

[0049] Specifically, for the noise-reduced signal s ( t Sparse wavelet transform (SSWT) is performed to obtain a time-frequency representation with highly concentrated energy.

[0050] 1. First, regarding s ( t Perform continuous wavelet transform (CWT) to obtain wavelet coefficients. W s ( a , b ):

[0051] 'a' is a scale factor, which is inversely proportional to the frequency. b This is the time shift factor, corresponding to the position on the time axis. ψ ( t ) is the mother function of a complex wavelet (such as the Morlet wavelet). To represent complex conjugation.

[0052] 2. Calculate the instantaneous frequency of the wavelet coefficients. ω s ( a , b ):

[0053] ω s (a,b) In order to scale a and time b The estimated instantaneous angular frequency (rad / s) at that point. j The imaginary unit, To find the partial derivative with respect to the time shift factor b.

[0054] 3. Synchronous compression operation: The wavelet transform coefficients are transferred from the scale-time plane (…). a , b "Squeezing" is mapped to the frequency-time plane. ω , b On the SSWT spectrum, T ( b , ω ):

[0055] T ( b , ω The time spectrum obtained after synchronous compression transformation is a complex numerical matrix. ω For angular frequency, in actual operation, for each ( a , b ), and its coefficient W s ( a , b Reassigned to frequency ω s ( a , b Superimpose on the corresponding positions.

[0056] The complex-time spectrum obtained by sparse wavelet transform T ( b , ω Converted to energy spectrum P ( t ,f ), and construct a three-dimensional map.

[0057]

[0058] P ( t m , f k ) for time t m and frequency f k Signal energy intensity at that location T m,k Spectrum of SSWT T ( b , ω At discrete time points m and discrete frequency points k The complex value at the position is represented by ||, which is the modulo operation.

[0059] In one embodiment, the frequency centroid coordinates in the three-dimensional energy are calculated using the following method:

[0060] in, P ( t m , f k ) for time t m and frequency f k Energy intensity at that location f c Let M be the frequency centroid coordinate in the three-dimensional energy, M be the total number of time discrete points, and K be the total number of frequency discrete points.

[0061] Specifically, the three-dimensional energy centroid features are extracted. The centroid coordinates of the three-dimensional energy distribution are calculated using the following formula to characterize the overall position of the signal energy in the time-frequency-energy space.

[0062]

[0063] ( t c , f c , P c () represents the three-dimensional energy barycentric coordinates. t c This represents the weighted average position of energy on the time axis. f c This refers to the weighted average position of energy on the frequency axis, i.e., the dominant frequency. PcThis is a weighted average of energy values, reflecting the degree of energy concentration. M is the total number of discrete points in time, and K... for The total number of frequency discrete points.

[0064] In one embodiment, calculating the three-dimensional energy entropy and band energy ratio based on a three-dimensional spectrum includes: Based on the three-dimensional map, the probability of the three-dimensional energy distribution is calculated, and the probability of the energy intensity distribution at the time and frequency points is obtained. The three-dimensional energy entropy is obtained by calculating the probability of energy intensity distribution at time and frequency points. Based on a preset frequency segmentation threshold, the time spectrum is divided into a first frequency band and a second frequency band. Based on the energy intensity in the three-dimensional spectrum, the total low-frequency energy of the first frequency band and the total high-frequency energy of the second frequency band are calculated respectively. The ratio of total high-frequency energy to total low-frequency energy is used as the frequency band energy ratio.

[0065] Specifically, the information entropy of the three-dimensional energy distribution is calculated to characterize the disorder and complexity of the signal energy.

[0066] 1. The three-dimensional energy distribution P ( t m , f k Probabilization:

[0067] p m,k For energy distribution at time and frequency points ( m , k The probability at position ).

[0068] 2. Calculate the three-dimensional energy entropy:

[0069] H 3D The value represents the three-dimensional energy entropy. The larger the value, the more dispersed and disordered the energy distribution; the smaller the value, the more concentrated and ordered the energy.

[0070] In the time-frequency plane P ( t , f On the energy peak, we can trace its evolution path.

[0071] 1. For each time slice t m Find its corresponding energy peak frequency f peak ( t m ):

[0072] f peak ( t m ) for time t m At that point, the frequency value with the strongest energy. Argmax is used to find the function... P ( t m , f k The one that gets the maximum value f k .

[0073] 2. Connect all ( t m , f peak ( t m The point forms the main energy trace.

[0074] To quantify the transfer of energy between high and low frequencies, the energy ratio of the high-frequency low-energy band to the low-frequency high-energy band is defined.

[0075] 1. Set frequency segmentation threshold f th The time spectrum is divided into the first frequency band (also called the low frequency band). f < f th ) and the first frequency band (also called the high frequency band) f ≥ f th ).

[0076] 2. Calculate the total energy for each of the two frequency bands:

[0077] E low This represents the total energy in the low-frequency band (which typically corresponds to macroscopic crack activity). E high This represents the total energy in the high-frequency band (which typically corresponds to microcrack activity). f th The threshold frequency for dividing the high and low frequency bands.

[0078] 3. Calculate the bandwidth energy ratio:

[0079] R HL This is the ratio of high-frequency energy to low-frequency energy. In the early stages of a rupture, this value often shows a trend of first rising and then falling.

[0080] In one embodiment, such as Figure 3 As shown, the smoothed frequency centroid coordinate feature sequence, energy entropy feature sequence, and frequency band energy ratio feature sequence are analyzed with their respective preset precursor feature sequences. Based on the analysis results, the warning level is obtained, including: 302: Calculate the first distance between the smoothed frequency centroid coordinate feature sequence and its corresponding preset precursor feature sequence, determine the trend of the frequency centroid coordinate within a preset time period based on the smoothed frequency centroid coordinate feature sequence, and determine the warning value of the frequency centroid coordinate based on the trend and the first distance; 304: Calculate the second distance between the smoothed energy entropy feature sequence and its corresponding preset precursor feature sequence, determine the energy entropy decrease mode based on the smoothed energy entropy feature sequence, and determine the energy entropy warning value based on the decrease mode and the second distance; 306: Calculate the third distance between the smoothed frequency band energy ratio feature sequence and its corresponding preset precursor feature sequence, determine the inflection point mode of the frequency band energy ratio based on the smoothed frequency band energy ratio feature sequence, and determine the warning value of the frequency band energy ratio according to the inflection point mode and the third distance; 308: The total warning level is calculated based on the warning values ​​of frequency centroid coordinates, energy entropy, and frequency band energy ratio, as well as their respective weights.

[0081] Specifically, the smoothed feature sequences monitored in real time Template sequences in a pre-established standard precursor feature library Perform dynamic time warping analysis to calculate the trend similarity distance. D DTW The following formula is used for calculation:

[0082] D DTW The dynamic time-normalized distance measures the similarity in shape between two time series; the smaller the value, the more similar they are. π For the optimal normalized path between two sequences, ( i , j () represents a pair of points on a normalized path, indicating The i Point and The j Point alignment.

[0083] Substituting the smoothed frequency centroid coordinate feature sequence and its corresponding preset precursor feature sequence into the above formula, the first distance is calculated. Substituting the smoothed energy entropy feature sequence and its corresponding preset precursor feature sequence into the above formula, the second distance is calculated. Substituting the smoothed frequency band energy ratio feature sequence and its corresponding preset precursor feature sequence into the above formula, the third distance is calculated. The trend of the frequency centroid coordinates within a preset time period is determined based on the smoothed frequency centroid coordinate feature sequence; the decreasing pattern of energy entropy is determined based on the smoothed energy entropy feature sequence; and the inflection point pattern of the frequency band energy ratio is determined based on the smoothed frequency band energy ratio feature sequence.

[0084] Establish a weighted voting fusion decision-making model. An early warning system of the corresponding level is triggered when multiple conditions are met in a coordinated change: 1. The trend within the preset time period is: When the short-term trend slope remains negative, and its first distance Below the threshold θ If 1, then the warning value for the frequency centroid coordinate feature is 1.

[0085] 2. The decreasing pattern of energy entropy is After reaching a high plateau, it experienced a sharp decline, and The warning value for the energy entropy characteristic is...

[0086] 3. Inflection point mode of bandwidth energy ratio An inflection point appeared, first rising and then falling, and The warning value for the frequency band energy ratio characteristic is 1.

[0087] Weights are assigned to each feature factor based on its historical early warning accuracy (frequency centroid coordinate feature, energy entropy feature, and bandwidth energy ratio feature). wi The final warning level will be determined by a weighted vote. (condition i (If true), among which I ( ) is an indicator function (1 if the condition is true, 0 otherwise), which calculates the total warning level.

[0088] In one implementation, an indoor test was used to simulate a complete coal and rock mass fracturing event to verify the ability of the method of the present invention to extract high-resolution time-frequency features from a single fracturing event signal, and to compare it with traditional methods (STFT, CWT).

[0089] The sparse wavelet transform method used in this invention generates a three-dimensional spectrum with extremely high energy concentration and significantly improved time-frequency resolution. This allows for more precise localization of the main energy distribution areas and instantaneous frequency changes, enabling the extracted feature indicators to more accurately reflect the physical nature of the signal and providing a reliable data foundation for accurate precursor identification. The original voltage signal used in this embodiment is as follows: Figure 2 As shown, the simulation performs real-time analysis of continuously generated microseismic / acoustic emission events in the actual underground coal mine environment, realizing early warning of rockburst precursors based on multi-feature synergistic evolution.

[0090] 1. Smooth the feature values ​​of 20 consecutive events to eliminate random fluctuations and obtain a smoothed sequence.

[0091] 2. Calculate the DTW distance between the current feature sequence and the "early signs of rockburst" template sequence in the pre-stored database in real time to measure their trend similarity.

[0092] 3. Perform multi-feature factor collaborative early warning; when processing the feature time series of the 16th event, Figure 4 As shown in the image.

[0093] 4. The weighted voting decision-maker considers the fulfillment of the above three conditions (all of which are met) and determines that the current state is a "precursor to a strong rockburst", immediately triggering the highest level of warning.

[0094] Furthermore, as a response to the above Figure 1 The implementation of the method shown in this invention provides a device for identifying precursor signals of coal and rock mass fracturing, such as... Figure 5 As shown, the device includes: The noise reduction module 502 is used to acquire the original voltage signal corresponding to the sound signal and / or vibration signal downhole, and to perform noise reduction processing on the original voltage signal to obtain the noise-reduced signal. The three-dimensional map construction module 504 is used to perform wavelet transform on the denoised signal to obtain the wavelet transform time spectrum, convert the wavelet transform time spectrum into an energy spectrum, and construct a three-dimensional map based on the wavelet transform time spectrum and the energy spectrum. The eigenvalue calculation module 506 is used to calculate the frequency centroid coordinates in the three-dimensional energy based on the three-dimensional spectrum, and to calculate the three-dimensional energy entropy and frequency band energy ratio based on the three-dimensional spectrum. The feature sequence acquisition module 508 is used to construct feature time series of frequency centroid coordinates, three-dimensional energy entropy and frequency band energy ratio in three-dimensional energy according to time order and perform smoothing processing to obtain smoothed frequency centroid coordinate feature series, energy entropy feature series and frequency band energy ratio feature series. The analysis module 510 is used to analyze the smoothed frequency centroid coordinate feature sequence, energy entropy feature sequence, and frequency band energy ratio feature sequence with their respective preset precursor feature sequences, and obtain the warning level based on the analysis results.

[0095] This application provides a device for identifying precursor signals of coal and rock mass fracturing. Compared with existing technologies, it reduces noise in the original voltage signal, performs time-frequency analysis and energy conversion on the noise-reduced signal, and constructs a high-resolution time-frequency-energy three-dimensional distribution map. Based on the three-dimensional distribution map, it extracts multi-dimensional feature indicators that can sensitively reflect the evolution process of coal and rock mass fracturing. It analyzes the multi-dimensional feature indicators and their corresponding preset precursor features, and obtains the warning level based on the analysis results. Through time-frequency analysis and time-frequency-energy three-dimensional joint feature extraction, it can capture weak, early precursor features that traditional methods cannot detect, significantly improving the identification accuracy and potentially obtaining a longer warning lead time. The noise reduction process can effectively distinguish between real fracturing signals and various underground environmental noises, greatly reducing the false alarm rate and missed alarm rate, and improving the accuracy and reliability of the warning.

[0096] In one embodiment, the noise reduction module is further used for: The original voltage signal is decomposed using an adaptive noise ensemble empirical mode decomposition method to obtain multiple intrinsic mode function components and residual terms. The effective signals of the intrinsic mode function components are then filtered to obtain effective component signals and invalid component signals. A semi-soft thresholding function is used to quantize the effective component signal and the invalid component signal respectively; The residual term, the quantized effective component signal, and the invalid component signal are reconstructed to obtain the denoised signal.

[0097] In one embodiment, the noise reduction module is further used for: Calculate the correlation coefficient and variance contribution rate between each intrinsic mode function component and the original voltage signal; The intrinsic mode function components with correlation coefficients greater than a preset coefficient threshold and variance contribution rates greater than a preset contribution rate are considered as valid component signals, while the remaining intrinsic mode function components are considered as invalid component signals.

[0098] In one embodiment, the three-dimensional mapping module is also used for: Perform continuous wavelet transform on the denoised signal to obtain wavelet coefficients, and calculate the instantaneous frequency of the wavelet coefficients; The wavelet transform spectrum is obtained by calculating the wavelet coefficients and their instantaneous frequencies, mapping them from the scale-time plane to the frequency-time plane. The square of the absolute value of the complex time spectrum in the wavelet transform time spectrum is taken as the energy spectrum corresponding to the wavelet transform time spectrum.

[0099] In one embodiment, the frequency centroid coordinates in the three-dimensional energy are calculated using the following method:

[0100] in, P ( t m , f k ) for time t m and frequency f k Energy intensity at that location f c Let M be the frequency centroid coordinate in the three-dimensional energy, M be the total number of time discrete points, and K be the total number of frequency discrete points.

[0101] In one embodiment, the eigenvalue calculation module is also used for: Based on the three-dimensional map, the probability of the three-dimensional energy distribution is calculated, and the probability of the energy intensity distribution at the time and frequency points is obtained. The three-dimensional energy entropy is obtained by calculating the probability of energy intensity distribution at time and frequency points. Based on a preset frequency segmentation threshold, the time spectrum is divided into a first frequency band and a second frequency band. Based on the energy intensity in the three-dimensional spectrum, the total low-frequency energy of the first frequency band and the total high-frequency energy of the second frequency band are calculated respectively. The ratio of total high-frequency energy to total low-frequency energy is used as the frequency band energy ratio.

[0102] In one embodiment, the analysis module is also used for: Calculate the first distance between the smoothed frequency centroid coordinate feature sequence and its corresponding preset precursor feature sequence, determine the trend of the frequency centroid coordinate within a preset time period based on the smoothed frequency centroid coordinate feature sequence, and determine the warning value of the frequency centroid coordinate based on the trend and the first distance. Calculate the second distance between the smoothed energy entropy feature sequence and its corresponding preset precursor feature sequence, determine the energy entropy decrease mode based on the smoothed energy entropy feature sequence, and determine the energy entropy warning value based on the decrease mode and the second distance. Calculate the third distance between the smoothed frequency band energy ratio feature sequence and its corresponding preset precursor feature sequence, determine the inflection point mode of the frequency band energy ratio based on the smoothed frequency band energy ratio feature sequence, and determine the warning value of the frequency band energy ratio based on the inflection point mode and the third distance. The total warning level is calculated based on the warning values ​​of frequency centroid coordinates, energy entropy, and frequency band energy ratio, as well as their respective weights.

[0103] According to one embodiment of the present invention, a storage medium is provided, the storage medium storing at least one executable instruction, which can execute the coal and rock mass fracturing precursor signal identification method in any of the above method embodiments.

[0104] Figure 6 The diagram illustrates a structural schematic of a computer device according to an embodiment of the present invention. The specific embodiments of the present invention do not limit the specific implementation of the computer device.

[0105] like Figure 6 As shown, the computer device may include: a processor 602, a communications interface 604, a memory 606, and a communications bus 608.

[0106] The processor 602, communication interface 604, and memory 606 communicate with each other via communication bus 608.

[0107] Communication interface 604 is used to communicate with other network elements such as clients or other servers.

[0108] The processor 602 is used to execute program 610, specifically to execute the relevant steps in the above-described embodiment of the method for identifying precursor signals of coal and rock mass fracturing.

[0109] Specifically, program 610 may include program code that includes computer operation instructions.

[0110] Processor 602 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention. The computer device includes one or more processors, which may be processors of the same type, such as one or more CPUs; or processors of different types, such as one or more CPUs and one or more ASICs.

[0111] Memory 606 is used to store program 610. Memory 606 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0112] Specifically, program 610 can be used to cause processor 602 to perform the following operations: Acquire the raw voltage signal corresponding to the sound signal and / or vibration signal downhole, and perform noise reduction processing on the raw voltage signal to obtain the noise-reduced signal; The denoised signal is subjected to wavelet transform to obtain the wavelet transform time spectrum. The wavelet transform time spectrum is then converted into an energy spectrum. Based on the wavelet transform time spectrum and energy spectrum, a three-dimensional spectrum is constructed. Based on the three-dimensional spectrum, calculate the frequency centroid coordinates in the three-dimensional energy, and calculate the three-dimensional energy entropy and frequency band energy ratio. The frequency centroid coordinates, three-dimensional energy entropy, and frequency band energy ratio in the three-dimensional energy are constructed into feature time series in chronological order and then smoothed to obtain smoothed frequency centroid coordinate feature series, energy entropy feature series, and frequency band energy ratio feature series. The smoothed frequency centroid coordinate feature sequence, energy entropy feature sequence, and frequency band energy ratio feature sequence are analyzed with their respective preset precursor feature sequences. Based on the analysis results, the warning level is obtained.

[0113] It will be apparent to those skilled in the art that the modules or steps of the present invention described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. In one embodiment, they can be implemented using device-executable program code, thereby allowing them to be stored in a storage device for execution by a computing device. In some cases, the steps shown or described can be performed in a different order than those presented herein, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the present invention is not limited to any particular hardware and software combination.

[0114] The above embodiments are merely exemplary embodiments of this application and are not intended to limit this application. The scope of protection of this application is defined by the claims. Those skilled in the art can make various modifications or equivalent substitutions to this application within its substance and scope of protection, and such modifications or equivalent substitutions should also be considered to fall within the scope of protection of this application.

Claims

1. A method for identifying precursor signals of coal and rock mass fracturing, characterized in that, include: Obtain the original voltage signal corresponding to the sound signal and / or vibration signal downhole, and perform noise reduction processing on the original voltage signal to obtain the noise-reduced signal; The denoised signal is subjected to wavelet transform to obtain the wavelet transform time spectrum, the wavelet transform time spectrum is converted into an energy spectrum, and a three-dimensional map is constructed based on the wavelet transform time spectrum and the energy spectrum. Based on the three-dimensional spectrum, calculate the frequency centroid coordinates in the three-dimensional energy, and calculate the three-dimensional energy entropy and frequency band energy ratio based on the three-dimensional spectrum; The frequency centroid coordinates, the three-dimensional energy entropy, and the frequency band energy ratio in the three-dimensional energy are respectively constructed into feature time series in time order and smoothed to obtain smoothed frequency centroid coordinate feature series, energy entropy feature series, and frequency band energy ratio feature series. The smoothed frequency centroid coordinate feature sequence, energy entropy feature sequence, and frequency band energy ratio feature sequence are analyzed with their respective preset precursor feature sequences. Based on the analysis results, the warning level is obtained.

2. The method for identifying precursor signals of coal and rock mass fracturing as described in claim 1, characterized in that, The step of denoising the original voltage signal to obtain the denoised signal includes: The original voltage signal is decomposed using an adaptive noise set empirical mode decomposition method to obtain multiple intrinsic mode function components and residual terms. The effective signals of the intrinsic mode function components are then filtered to obtain effective component signals and invalid component signals. The effective component signal and the invalid component signal are quantized using a semi-soft thresholding function respectively; The residual term, the quantized effective component signal, and the invalid component signal are reconstructed to obtain the denoised signal.

3. The method for identifying precursor signals of coal and rock mass fracturing as described in claim 2, characterized in that, The process of filtering the intrinsic mode function components to obtain valid and invalid component signals includes: Calculate the correlation coefficient and variance contribution rate between each intrinsic mode function component and the original voltage signal; The intrinsic mode function components with correlation coefficients greater than a preset coefficient threshold and variance contribution rates greater than a preset contribution rate are considered as valid component signals, while the remaining intrinsic mode function components are considered as invalid component signals.

4. The method for identifying precursor signals of coal and rock mass fracturing as described in claim 2, characterized in that, The step of performing wavelet transform on the denoised signal to obtain the wavelet transform time spectrum, and converting the wavelet transform time spectrum into an energy spectrum includes: Perform continuous wavelet transform on the denoised signal to obtain wavelet coefficients, and calculate the instantaneous frequency of the wavelet coefficients; The wavelet coefficients and their instantaneous frequencies are calculated, and the wavelet coefficients are mapped from the scale-time plane to the frequency-time plane to obtain the wavelet transform spectrum. The square of the absolute value of the complex time spectrum in the wavelet transform time spectrum is taken as the energy spectrum corresponding to the wavelet transform time spectrum.

5. The method for identifying precursor signals of coal and rock mass fracturing as described in claim 1, characterized in that, The frequency centroid coordinates in the three-dimensional energy are calculated using the following method: in, P ( t m , f k ) for time t m and frequency f k Energy intensity at that location f c Let M be the frequency centroid coordinate in the three-dimensional energy, M be the total number of time discrete points, and K be the total number of frequency discrete points.

6. The method for identifying precursor signals of coal and rock mass fracturing as described in claim 1, characterized in that, The calculation of the three-dimensional energy entropy and band energy ratio based on the three-dimensional spectrum includes: Based on the three-dimensional map, the probability of the three-dimensional energy distribution is calculated to obtain the probability of the energy intensity distribution at the time and frequency points; The three-dimensional energy entropy is obtained by calculating the probability of the energy intensity distribution at time and frequency points. Based on a preset frequency segmentation threshold, the time spectrum is divided into a first frequency band and a second frequency band. Based on the energy intensity in the three-dimensional spectrum, the total low-frequency energy of the first frequency band and the total high-frequency energy of the second frequency band are calculated respectively. The ratio of the total high-frequency energy to the total low-frequency energy is taken as the frequency band energy ratio.

7. The method for identifying precursor signals of coal and rock mass fracturing as described in claim 1, characterized in that, The process involves analyzing the smoothed frequency centroid coordinate feature sequence, energy entropy feature sequence, and frequency band energy ratio feature sequence with their respective preset precursor feature sequences. Based on the analysis results, an early warning level is obtained, including: Calculate the first distance between the smoothed frequency centroid coordinate feature sequence and its corresponding preset precursor feature sequence, determine the trend of the frequency centroid coordinate within a preset time period based on the smoothed frequency centroid coordinate feature sequence, and determine the warning value of the frequency centroid coordinate based on the trend and the first distance. Calculate the second distance between the smoothed energy entropy feature sequence and its corresponding preset precursor feature sequence, determine the energy entropy decrease mode based on the smoothed energy entropy feature sequence, and determine the energy entropy warning value based on the decrease mode and the second distance; Calculate the third distance between the smoothed frequency band energy ratio feature sequence and its corresponding preset precursor feature sequence, determine the inflection point mode of the frequency band energy ratio based on the smoothed frequency band energy ratio feature sequence, and determine the warning value of the frequency band energy ratio according to the inflection point mode and the third distance. The total warning level is calculated based on the warning values ​​of the frequency centroid coordinates, the energy entropy, and the frequency band energy ratio, as well as their respective weights.

8. A device for identifying precursor signals of coal and rock mass fracturing, characterized in that, include: The noise reduction module is used to acquire the original voltage signal corresponding to the sound signal and / or vibration signal downhole, and to perform noise reduction processing on the original voltage signal to obtain the noise-reduced signal. A three-dimensional atlas construction module is used to perform wavelet transform on the denoised signal to obtain the wavelet transform time spectrum, convert the wavelet transform time spectrum into an energy spectrum, and construct a three-dimensional atlas based on the wavelet transform time spectrum and the energy spectrum; The eigenvalue calculation module is used to calculate the frequency centroid coordinates in the three-dimensional energy based on the three-dimensional spectrum, and to calculate the three-dimensional energy entropy and frequency band energy ratio based on the three-dimensional spectrum. The feature sequence acquisition module is used to construct feature time series of the frequency centroid coordinates, the three-dimensional energy entropy and the frequency band energy ratio in the three-dimensional energy according to the time sequence and perform smoothing processing to obtain smoothed frequency centroid coordinate feature series, energy entropy feature series and frequency band energy ratio feature series. The analysis module is used to analyze the smoothed frequency centroid coordinate feature sequence, energy entropy feature sequence, and frequency band energy ratio feature sequence with their respective preset precursor feature sequences, and obtain the warning level based on the analysis results.

9. A storage medium storing at least one executable instruction that causes a processor to perform an operation corresponding to the coal and rock mass fracturing precursor signal identification method as described in any one of claims 1-7.

10. A computer device, comprising: The processor, memory, communication interface, and communication bus are provided, wherein the processor, memory, and communication interface communicate with each other via the communication bus. The memory is used to store at least one executable instruction, which causes the processor to perform the operation corresponding to the coal and rock mass fracturing precursor signal identification method as described in any one of claims 1-7.

Images