Coal rock charge signal energy calculation and rock burst monitoring and early warning method

Abstract

The present application provides a kind of coal rock charge signal energy calculation and rock burst monitoring and early warning method, it is related to mine dynamic disaster monitoring technical field.The method first obtains the charge signal of the deformation and failure process of loaded coal rock, and carries out joint denoising;Wavelet decomposition and reconstruction are carried out, and decomposition is carried out using Hilbert transformation, and the analytical signal of charge signal based on wavelet decomposition is constructed;The instantaneous amplitude of analytical signal is remapped in frequency to obtain the time-frequency energy distribution function of charge signal;Further calculate the time-frequency energy density function of charge signal, and then determine the charge signal energy of each measuring point;Then the non-parametric empirical quantile method is used to divide the charge signal energy, and the impact risk level corresponding to the charge signal energy level of each measuring point is determined;Finally, the comprehensive impact risk level is determined by using entropy weight method combined with the charge signal energy and stress corresponding impact risk level.

Description

Technical Field

[0001] This invention relates to the field of mine dynamic disaster monitoring technology, and in particular to a method for calculating the energy of coal and rock charge signals and for monitoring and early warning of rockburst. Background Technology

[0002] With the continuous increase in the depth and intensity of coal mining, rockburst has become a key dynamic disaster restricting the safe and efficient production of deep mines. During the deformation and fracturing process of loaded coal and rock, charge-induced signals are generated. These signals, due to their advantages such as high sensitivity, obvious precursor characteristics, and the ability to be monitored non-contactly, have gradually become a research hotspot in the field of rockburst monitoring and early warning. Currently, scholars at home and abroad have conducted extensive research on the generation mechanism, signal characteristics, and engineering applications of coal and rock charge signals. This research has confirmed the existence of charge separation during the deformation and failure process of coal and rock, and that this phenomenon is the core prerequisite for the generation of coal and rock charge signals, laying the foundation for subsequent monitoring and application of charge signals. Throughout the entire process from loading and deformation to instability and failure of coal and rock, there is always the accumulation, transformation, and dissipation of elastic strain energy. The generation and release of charge signals are essentially important manifestations of the dissipation of strain energy within the coal and rock. The strength and duration of the charge signal are directly related to the rate and total amount of strain energy dissipation. Charge signal energy, as a core parameter characterizing strain energy dissipation, can more accurately reflect the internal damage evolution process and impact hazard level of coal and rock, making it a key indicator for rockburst early warning. Although the intrinsic relationship between charge signal and coal and rock strain energy dissipation is clear, and charge signal energy is significant for rockburst hazard early warning, methods for calculating charge signal energy remain lacking. Existing research mostly focuses on the statistical level of basic parameters such as charge signal amplitude and event rate, without establishing a quantitative correlation between charge signal energy and the accumulation-release of elastic strain energy in coal and rock, and the degree of damage evolution. Regarding the determination of impact hazard warning thresholds, the core shortcoming of existing research lies in the lack of thresholds related to charge signal energy. Because a method for calculating charge signal energy has not yet been established, the industry has not yet formed an impact hazard warning threshold based on charge signal energy. There is a lack of scientific definition of the critical value of charge signal energy, and no multi-dimensional, dynamic, hierarchical threshold system based on charge signal energy has been constructed. This makes it impossible to achieve accurate early warning of rockbursts using charge signal energy as a core indicator, which has become a major obstacle restricting the advancement of charge monitoring technology towards precise and engineering applications. Furthermore, during field monitoring, charge signals are easily affected by factors such as electromagnetic interference from mechanical equipment, coal and rock heterogeneity, and dynamic stress fluctuations. Even if a method for calculating charge signal energy is subsequently established, key issues such as interference signal removal and energy characteristic normalization need to be addressed. Existing methods have not yet formed effective solutions, further exacerbating the difficulties in applying charge signal energy in impact early warning applications. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide a method for calculating the energy of coal and rock charge signals and for monitoring and early warning of rockbursts, which addresses the shortcomings of the prior art. This method obtains the evolution law of charge signal energy during the deformation and failure of coal and rock, and clarifies the graded early warning of rockburst hazards.

[0004] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: a method for calculating the energy of coal and rock charge signals and for monitoring and early warning of rockbursts, comprising the following steps:

[0005] Step 1: Set up multiple measuring points to acquire charge signals during the deformation and failure process of loaded coal and rock;

[0006] Step 2: Perform FFT band-stop filtering and wavelet transform combined to denoise the charge signals acquired at each measurement point;

[0007] First, power frequency noise in the 50 Hz ± 0.5 Hz frequency band and its harmonics is removed by Fourier transform. Then, non-target transient components that are approximately randomly distributed in the wavelet scale domain are removed by discrete wavelet transform. Among them, the db8 wavelet is selected as the basis function for discrete wavelet transform, and the adaptive threshold and hard threshold functions of Stein unbiased risk estimation threshold are used to process the charge signal after power frequency noise removal, and finally the effective charge signal after joint denoising is obtained.

[0008] Step 3: Perform a second wavelet decomposition and reconstruction on the jointly denoised charge signal to obtain wavelet reconstruction components at different scales;

[0009] The charge signal after joint denoising is decomposed using J-layer discrete wavelets, and then the wavelet coefficients at each scale are reconstructed layer by layer to obtain wavelet reconstruction components at different scales.

[0010] Step 4: Use Hilbert transform to decompose the wavelet reconstruction components at different scales and construct the analytic signal of the charge signal based on wavelet decomposition;

[0011] The constructed charge signal is an analytic signal based on wavelet decomposition. for:

[0012] ;

[0013] In the formula: The k-th wavelet reconstructed component of the jointly denoised charge signal. for Hibert transform, for The instantaneous amplitude; for The instantaneous phase; i is the imaginary part, satisfying... ;

[0014] Step 5: Analyze the signal The instantaneous amplitude is remapped in frequency to obtain the time-frequency energy distribution function of the charge signal. As shown in the formula below:

[0015] ;

[0016] In the formula: for The instantaneous frequency; Indicates the frequency of the charge signal; For Dirac Distribution, used to represent the instantaneous amplitude of the k-th component. Positioned at its corresponding instantaneous frequency ;

[0017] Step 6: Time-frequency energy distribution function of charge signal Based on this, the time-frequency energy density function of the charge signal was further calculated. As shown in the formula below:

[0018] ;

[0019] Step 7: Determine the charge signal energy at each measurement point based on the time-frequency energy density function of the charge signal;

[0020] Time-frequency energy density function The integral along the frequency direction is defined as the charge signal energy of the charge signal. As shown in the formula below:

[0021] ;

[0022] Step 8: Use the non-parametric empirical quantile method to classify the charge signal energy and determine the impact hazard level corresponding to different charge signal energy levels at each measuring point;

[0023] Calculate the 90%, 95%, and 99% empirical quantiles of the charge signal energy during the entire loading process of the coal body, denoted as . , , Based on this, the impact hazard classification thresholds are determined, and then the impact hazard level L corresponding to different charge energy levels at each measuring point is determined. E ;when The corresponding impact hazard level L E =0, meaning there is no risk of impact. The corresponding impact hazard level L E =1, meaning a weak impact hazard, when The corresponding impact hazard level L E =2, meaning moderate impact risk, when The corresponding impact hazard level L E =3, indicating a high risk of impact;

[0024] Step 9: Determine the impact hazard level L corresponding to the charge signal energy level at each measuring point using the entropy weight method. E The impact hazard level L corresponding to the stress level near the charge measurement point K The weights are used to determine the overall impact hazard level L. C ;

[0025] First, the charge signal energy corresponding to each measuring point is normalized, and the proportion and entropy value of the charge signal energy and stress corresponding to each measuring point are calculated. Then, the entropy redundancy of the charge signal energy and stress is calculated. Based on the entropy redundancy of the charge signal energy and stress, the impact hazard level L based on the charge signal energy is calculated. E Impact hazard level L corresponding to stress level K weight and Finally, the comprehensive impact hazard level L was obtained. C :

[0026] ;

[0027] In the formula: The impact hazard level determined at any given time at this measuring point based on the charge signal energy; L K This represents the impact hazard level corresponding to the stress level near the charge measurement point.

[0028] The beneficial effects of adopting the above technical solution are as follows: The coal and rock charge signal energy calculation and rockburst monitoring and early warning method provided by this invention integrates wavelet decomposition and Hilbert transform methods, defines and proposes a charge signal energy index, and provides a new method for indirectly and quantitatively characterizing the energy dissipation process of coal. The 90%, 95%, and 99% empirical quantiles of the charge signal energy are selected to reasonably divide the rockburst hazard classification threshold, thereby determining the rockburst hazard level corresponding to the charge signal energy level. Combining the entropy weight method, the rockburst hazard level based on the charge signal and the rockburst hazard level based on the stress level are fused to construct a comprehensive rockburst hazard level index. The constructed comprehensive rockburst hazard level index can effectively reflect the dangerous state of the coal face. Attached Figure Description

[0029] Figure 1 A flowchart illustrating a method for calculating coal and rock charge signal energy and monitoring and early warning of rockburst, provided in an embodiment of the present invention;

[0030] Figure 2This is a schematic diagram of the impact hazard classification threshold based on charge signal energy division for a tunneling face in a coal mine in Ningxia, provided by an embodiment of the present invention.

[0031] Figure 3 This is a schematic diagram illustrating the classification of the comprehensive impact hazard level of a tunneling face in a mine in Ningxia, based on the entropy weight method, as provided in an embodiment of the present invention. Detailed Implementation

[0032] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.

[0033] In this embodiment, a method for calculating the energy of coal and rock charge signals and for monitoring and early warning of rockbursts is described, such as... Figure 1 As shown, it includes the following steps:

[0034] Step 1: Set up multiple measuring points to acquire charge signals during the deformation and failure process of loaded coal and rock;

[0035] Step 2: Perform FFT band-stop filtering and wavelet transform combined to denoise the charge signals acquired at each measurement point;

[0036] To obtain an accurate effective charge signal, a joint denoising method is required: First, power frequency noise in the 50 Hz ± 0.5 Hz band and its harmonics is removed by Fourier transform. Then, non-target transient components that are approximately randomly distributed in the wavelet scale domain are removed by discrete wavelet transform. Specifically, the db8 wavelet is selected as the basis function for discrete wavelet transform, and the adaptive threshold and hard threshold functions of Stein's unbiased risk estimation threshold (SURE) are used to process the charge signal after power frequency noise removal, finally obtaining the effective charge signal after joint denoising.

[0037] The charge signal after power frequency noise has been removed is denoted as The charge signal is decomposed into wavelet coefficients at a specified number of levels. As shown in the formula below:

[0038] (1);

[0039] In the formula: is the wavelet basis function; in this embodiment, the db8 wavelet is selected; s is the scaling factor; b is the translation factor; and t is the time.

[0040] The Stein unbiased risk estimation threshold is calculated by sorting the squares of the wavelet coefficients from smallest to largest to obtain a new vector. Calculate the risk value for:

[0041] (2);

[0042] In the formula: l represents the sequence number of the squared values ​​of the wavelet coefficients arranged in ascending order, and ; Assumption hour, Risk value The minimum value to be determined is used to locate the position. and the corresponding wavelet coefficient values ;

[0043] Then adaptive threshold Choose from the following:

[0044] (3);

[0045] In the formula: This is an estimate of the standard deviation of the noise. The median of the wavelet coefficient amplitudes, sorted by magnitude.

[0046] The hard threshold function is:

[0047] (4);

[0048] In the formula: The wavelet coefficients obtained from the j-th layer of wavelet decomposition before thresholding. This represents the wavelet coefficients retained after thresholding. The threshold for the j-th layer is determined by wavelet decomposition.

[0049] Step 3: Perform a second wavelet decomposition and reconstruction on the jointly denoised charge signal to obtain wavelet reconstruction components at different scales;

[0050] The charge signal after joint denoising Using J-layer discrete wavelet decomposition, we can obtain:

[0051] (5);

[0052] In the formula: For the wavelet reconstruction detail components of the j-th layer of the jointly denoised charge signal, ; For the wavelet reconstruction approximation components of the jointly denoised charge signal;

[0053] The wavelet coefficients at each scale are then reconstructed layer by layer in reverse to obtain wavelet reconstruction components at different scales, thus achieving multi-scale component separation of the jointly denoised charge signal. In this embodiment, the wavelet reconstruction components at each scale of the jointly denoised charge signal are... As shown in the formula below:

[0054] (6);

[0055] Step 4: Use Hilbert transform to decompose wavelet reconstruction components at different scales and construct an analytic signal based on wavelet decomposition;

[0056] Wavelet reconstructed components and Perform convolution operations, that is, reconstruct the wavelet components at each scale. The Hibert transform is shown in the following formula:

[0057] (7);

[0058] In the formula: for Hibert transform, Let P be the integral variable, and let P be the Cauchy principal value, used to emphasize the locality of the signal.

[0059] Further construct the analytic signal of the charge signal based on wavelet decomposition. for:

[0060] (8);

[0061] In the formula: for The instantaneous amplitude; for The instantaneous phase; i is the imaginary part, satisfying... .

[0062] Step 5: Considering that the instantaneous frequency is the first derivative of the instantaneous phase with respect to time, therefore, the analytic signal of the charge signal based on wavelet decomposition... The instantaneous amplitude is remapped in frequency to obtain the time-frequency energy distribution function of the charge signal. As shown in the formula below:

[0063] (9);

[0064] In the formula: for The instantaneous frequency; Indicates the frequency of the charge signal. For Dirac Distribution, used to represent the instantaneous amplitude of the k-th component. Positioned at its corresponding instantaneous frequency .

[0065] Step 6: Time-frequency energy distribution function of charge signal Based on this, the time-frequency energy density function of the charge signal was further calculated. As shown in the formula below:

[0066] (10);

[0067] Step 7: Determine the charge signal energy at each measurement point based on the time-frequency energy density function of the charge signal;

[0068] The time-frequency energy density function of the charge signal The integral along the frequency direction is defined as the charge signal. Charge signal energy As shown in the formula below:

[0069] (11);

[0070] Step 8: Use the non-parametric empirical quantile method to classify the charge signal energy and determine the impact hazard level corresponding to different charge signal energy levels at each measuring point;

[0071] Calculate the 90%, 95%, and 99% empirical quantiles of the charge signal energy during the entire loading process of the coal body, denoted as . , , Based on this, the impact hazard classification thresholds are determined, and then the impact hazard level L corresponding to the charge energy level at each measuring point is determined. E ;when The corresponding impact hazard level L E =0, meaning there is no risk of impact. The corresponding impact hazard level L E =1, meaning a weak impact hazard, when The corresponding impact hazard level L E =2, meaning moderate impact risk, when The corresponding impact hazard level L E =3, indicating a high risk of impact;

[0072] Table 1 Impact Hazard Levels Based on Charge Signal Energy

[0073]

[0074] This invention employs a nonparametric empirical quantile method to calculate the energy of coal and rock charge signals collected at each measuring point and classify their corresponding impact hazard levels. The entire process does not rely on normal distribution, log-normal distribution, or other preset probability distribution models, nor does it require distribution testing or fitting of the original charge signals. It is entirely based on the adaptive determination of impact hazard levels using charge signal energy samples.

[0075] Step 9: Determine the impact hazard level L corresponding to the charge signal energy level at each measuring point using the entropy weight method. E The impact hazard level L corresponding to the stress level near the charge measurement point K The weights are used to determine the overall impact hazard level L. C ;

[0076] Ground-based field and experimental data show that charge signals are highly sensitive to local microcracks and damage in coal and rock, and relying solely on charge signal energy may lead to inaccurate hazard assessment. To overcome this problem in routine field monitoring, this invention introduces an impact hazard level L corresponding to the stress level near each charge measurement point. K Impact hazard level L corresponding to the energy level of the charge signal E The overall impact hazard level was determined by integration.

[0077] Impact hazard level L corresponding to the stress level near each charge measuring point K The values ​​are also 0, 1, 2, and 3, which correspond to no impact risk, weak impact risk, moderate impact risk, and strong impact risk, respectively.

[0078] First, the charge signal energy and stress corresponding to each measuring point are normalized:

[0079] (12);

[0080] In the formula: , The normalized charge signal energy at any given time; , These are the minimum and maximum values ​​of the charge signal energy, respectively; , These represent the minimum and maximum values ​​of stress, respectively.

[0081] Next, the ratio of charge signal energy to stress and the entropy value are calculated:

[0082] (13);

[0083] In the formula: This represents the proportion of the normalized charge signal energy at this measurement point in the total normalized charge signal energy at all measurement points. This represents the proportion of the normalized stress at this measuring point to the total normalized stress at all measuring points.

[0084] (14);

[0085] In the formula: This represents the entropy value corresponding to the charge signal energy; m is the total number of charge signal energy samples corresponding to this charge measurement point; This represents the entropy value corresponding to the stress; n is the total number of stress samples.

[0086] Further calculation of the entropy redundancy of charge signal energy and stress. , The following is a public notice:

[0087] (15);

[0088] Since the stress value remains constant during charge monitoring, the entropy redundancy of the impact hazard level index corresponding to the stress level near the charge measuring point is set to 1. Therefore, the weights of the two indices are as follows:

[0089] (16);

[0090] In the formula: and The impact hazard level L corresponding to the charge signal energy level determined by the entropy weight method. E Impact hazard level L corresponding to stress level K The weights;

[0091] Finally, the comprehensive impact hazard level L is obtained based on the charge signal energy hazard level and the corresponding hazard level of stress. C :

[0092] (17);

[0093] In the formula: L represents the impact hazard level corresponding to the charge signal energy level at any given time at this measuring point. K This represents the impact hazard level corresponding to the stress level near the charge measurement point.

[0094] In this embodiment, the comprehensive impact hazard level range is defined as follows: L C ∈[0,0.3) indicates no impact risk, L C ∈(0.3,0.5] represents a weak impact hazard, L C ∈(0.5,0.7] represents a moderate impact hazard, L C The area ∈(0.7,1] represents a strong impact hazard.

[0095] This embodiment uses charge signal data acquired from nine measuring points at a tunneling face in a mine in Ningxia as an example. First, joint denoising is performed on the original charge signal. Then, using the charge signal energy calculation method of this invention, a charge signal energy sequence is obtained. Based on this energy sequence, a non-parametric empirical quantile method is used to calculate the 90%, 95%, and 99% quantiles as grading thresholds, such as... Figure 2 As shown.

[0096] In this embodiment, the 90th percentile of the acquired charge signal energy sequence is: (Lower risk threshold), 95th percentile is (Lower threshold for moderate risk), 99th percentile is (Lower threshold of high risk).

[0097] A comprehensive analysis of the impact hazard level was then conducted by combining monitoring data from a tunneling face in a mine in Ningxia. Since the stress near each measuring point on the working face was less than 5 MPa, the impact hazard level L corresponding to this stress level was determined. K The value is set to 0. Based on the entropy weight method, the comprehensive impact hazard level index corresponding to different measuring points can be obtained, such as... Figure 3 As shown, the results are consistent with the actual situation, and the comprehensive impact hazard level index can be applied to on-site monitoring and early warning.

[0098] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope defined by the present invention.

Claims

1. A method for calculating the energy of coal and rock charge signals and for monitoring and early warning of rockbursts, comprising the following steps: Step 1: Set up multiple measuring points to acquire charge signals during the deformation and failure process of loaded coal and rock; Step 2: Perform FFT band-stop filtering and wavelet transform combined to denoise the charge signals acquired at each measurement point; Step 3: Perform a second wavelet decomposition and reconstruction on the jointly denoised charge signal to obtain wavelet reconstruction components at different scales; Step 4: Use Hilbert transform to decompose the wavelet reconstruction components at different scales and construct the analytic signal of the charge signal based on wavelet decomposition; The constructed charge signal is based on the analytic signal of the wavelet decomposition For: ； wherein: is the kth wavelet reconstruction component of the denoised charge signal, is the kth wavelet reconstruction component of the denoised charge signal, is the Hibert transform of is the kth wavelet reconstruction component of the denoised charge signal, is the instantaneous amplitude of is the instantaneous phase of is the instantaneous phase of i For imaginary units to satisfy ; Step 5: Remap the instantaneous amplitude of the analytical signal in terms of frequency to obtain the time-frequency energy distribution function of the charge signal; The time-frequency energy distribution function of the charge signal is shown in the following formula: ； In the formula: Let be the time-frequency energy distribution function of the charge signal. for The instantaneous frequency; Indicates the frequency of the charge signal; For Dirac Distribution, used to divide the first k Instantaneous amplitude of each wavelet reconstructed component Positioned at its corresponding instantaneous frequency ; Step 6: Based on the time-frequency energy distribution function of the charge signal, further calculate the time-frequency energy density function of the charge signal; The time-frequency energy density function of the charge signal is shown in the following formula: ； in, Let be the time-frequency energy density function of the charge signal; Step 7: Determine the charge signal energy at each measurement point based on the time-frequency energy density function of the charge signal; The integral of the time-frequency energy density function along the frequency direction is defined as the charge signal energy of the charge signal. As shown in the formula below: ； Step 8: Use a non-parametric empirical quantile method to classify the charge signal energy and determine the impact hazard level corresponding to the charge signal energy level at each measuring point. L E ; Calculate the 90%, 95%, and 99% empirical quantiles of the charge signal energy during the entire loading process of the coal body, denoted as . , , Based on this, the impact hazard classification threshold is determined, thereby identifying the impact hazard level corresponding to the charge energy level. L E ;when Corresponding impact hazard level L E =0, meaning there is no risk of impact. Corresponding impact hazard level L E =1, meaning a weak impact hazard, when Corresponding impact hazard level L E =2, meaning moderate impact risk, when Corresponding impact hazard level L E =3, indicating a high risk of impact; Step 9: Determine the impact hazard level corresponding to the charge signal energy level at each measuring point using the entropy weight method. L E Impact hazard level corresponding to the stress level near the measuring point L K The weights are used to determine the overall impact risk level. L C ; First, the charge signal energy corresponding to each measuring point is normalized, and the proportion and entropy value of charge signal energy and stress at each measuring point are calculated. Then, the entropy redundancy of charge signal energy and stress is calculated. Based on the entropy redundancy of charge signal energy and the entropy redundancy corresponding to stress, the impact hazard level based on charge signal energy is calculated. L E Impact hazard level corresponding to stress level L K weight and Finally, the comprehensive impact hazard level was obtained. L C : ； In the formula: The impact hazard level is determined based on the charge signal energy at any given time at this measuring point; L K This represents the impact hazard level corresponding to the stress level near the measuring point.

2. The method for calculating coal and rock charge signal energy and monitoring and early warning of rockburst according to claim 1, characterized in that, The specific method for step 2 is as follows: First, power frequency noise in the 50 Hz ± 0.5 Hz frequency band and its harmonics is removed by Fourier transform. Then, discrete wavelet transform is used to remove non-target transient components that are approximately randomly distributed in the wavelet scale domain. Specifically, the discrete wavelet transform uses the db8 wavelet as the basis function and employs the adaptive threshold and hard threshold functions of Stein's unbiased risk estimation threshold to process the charge signal after power frequency noise removal, finally obtaining the effective charge signal after joint denoising.

3. The method for calculating coal and rock charge signal energy and monitoring and early warning of rockburst according to claim 2, characterized in that, Step 3 uses the charge signal after joint denoising. J The discrete wavelet is decomposed layer by layer, and then the wavelet coefficients at each scale are reconstructed layer by layer to obtain wavelet reconstruction components at different scales.

Images