Mining-induced seismic source inversion method and system

Abstract

This invention discloses a method and system for inverting seismic sources in deep coal mines, relating to the field of deep coal mining safety technology. The seismic source inversion method includes the following steps: acquiring initial seismic signal data; processing the data using the Fast Fourier Transform-Butterworth method to obtain seismic signal waveform data; extracting the seismic signal waveform data to form a microseismic signal database, and adding the seismic occurrence date, time, and seismic sensor number to each signal data in the microseismic signal database; extracting the first arrival time of each signal data in the microseismic signal database using the sliding window energy ratio method; and locating the actual seismic source coordinates using a genetic algorithm. Thus, by using the Fast Fourier Transform-Butterworth method and the genetic algorithm, high-precision actual seismic source coordinates can be obtained, making the seismic source location error within 5%.

Description

Technical Field

[0001] This invention relates to the field of deep coal mine safety mining technology, specifically to a method and system for inverting mine seismic sources. Background Technology

[0002] Coal is my country's primary energy source and an important industrial raw material. In recent years, frequent strong mine tremors during deep coal mining have severely impacted the safe and efficient production of mines. Currently, research on mine tremors still faces challenges such as "unclear mine tremor mechanisms, inaccurate source location, and unclear pattern prediction."

[0003] In seismic monitoring technology, source location is a key factor determining its reliability, and the accuracy of source location will have a certain impact on subsequent data interpretation. Source location refers to determining the spatial coordinates and time of the seismic source, and it is the most classic and fundamental problem in microseismic research. The basic approach to source location is based on iterative forward and inverse model inversion, using the picked first arrivals of direct P and S waves to invert the source location or time of origin, minimizing the error between the simulated and actual first arrivals. This method is currently the most widely used. In recent years, more and more new methods have been proposed, and various optimization algorithms have been applied to source location. However, current source location inversion methods still suffer from problems such as low accuracy of source inversion parameters. Summary of the Invention

[0004] The main objective of this invention is to propose a method and system for inverting the seismic source of a mine earthquake with high positioning accuracy.

[0005] To achieve the above objectives, this invention proposes a method for inverting the source of a mine-induced seismic event, which includes the following steps:

[0006] Step S1: Obtain initial seismic signal data;

[0007] Step S2: Process the data using the Fast Fourier Transform-Butterworth method to obtain the waveform data of the mine seismic signal;

[0008] Step S3: Extract the waveform data of the seismic signal to form a microseismic signal database, and add the date, time and sensor number of the seismic event to each signal data in the microseismic signal database;

[0009] Step S4: Extract the first arrival time for each signal data in the microseismic signal database using the sliding window energy ratio method;

[0010] Step S5: Use a genetic algorithm to locate the actual seismic source coordinates.

[0011] Further, step S2 specifically includes:

[0012] Step S21: Normalize the initial seismic signal data to obtain normalized signal data;

[0013] Step S22: Perform a fast Fourier transform on the normalized signal data to obtain the single-sided amplitude spectrum and frequency domain information of the seismic signal;

[0014] Step S23: Use a Butterworth bandpass filter to filter and reduce noise in the normalized signal data to obtain the signal waveform diagram and the mine seismic signal waveform data.

[0015] Furthermore, step S3 specifically includes:

[0016] Step S31: Extract clear actual seismic signal data from the normalized signal data, perform noise reduction processing, and integrate them to form the microseismic signal database;

[0017] Step S32: Add the date and time of the mine earthquake and the mine earthquake sensor number to each signal data in the microseismic signal database.

[0018] Further, step S31 specifically includes:

[0019] Step S311: Based on the single-sided amplitude spectrum, extract clear actual mine seismic signal data from the normalized signal data and perform noise reduction processing;

[0020] Step S312: Based on the signal waveform diagram, and considering the background noise frequency range and signal-to-noise ratio, remove signal data from the actual mine seismic signal data from which noise and mine seismic waveforms cannot be separated, thereby forming the microseismic signal database.

[0021] Further, step S1 specifically includes:

[0022] The initial seismic signal data is acquired using a coal mine microseismic monitoring system, wherein the coal mine microseismic monitoring system includes multiple sensors;

[0023] Step S5 specifically includes:

[0024] Step S51: Obtain the corresponding seismic source coordinates based on the position coordinates of each sensor, and confirm the transmission speed of the seismic wave.

[0025] Step S52: Use the genetic algorithm to invert and obtain the coordinates of the actual seismic source.

[0026] Further, step S51 specifically includes:

[0027] Step S511: Obtain the position coordinates of each sensor;

[0028] Step S512: Obtain the position coordinates of each sensor according to the positioning equation to obtain the corresponding seismic source position coordinates, and confirm the transmission velocity of the seismic wave. The positioning equation is as follows: Where (x0, y0, z0) are the coordinates of the earthquake source location, (x i ,y i ,z i Let t0 be the position coordinates of the i-th sensor, i = 1, 2, 3, ..., n, and t0 be the start time of the mine tremor. i denoted as , where is the time it takes for the seismic wave to reach the sensor; v is the transmission speed of the seismic wave.

[0029] Further, step S52 specifically includes:

[0030] Step S521: Initialize the population;

[0031] Step S522: Initialize the coordinates of the earthquake source location;

[0032] Step S523: Fitness of each individual in the population;

[0033] Step S524: Select the individual with the highest fitness and perform crossover, mutation, and genetics.

[0034] Step S525: Update the population;

[0035] Step S526: Repeat steps S522-S525 until the maximum number of iterations is reached or a certain threshold is met to obtain the actual seismic source coordinates.

[0036] Furthermore, after step S5, the method further includes:

[0037] Step S6: Perform error analysis on the actual seismic source coordinates.

[0038] This invention also provides a mine-induced seismic source inversion system, applicable to mine-induced seismic source inversion methods, the mine-induced seismic source inversion system comprising:

[0039] The seismic signal acquisition module is used to acquire initial seismic signal data;

[0040] The signal processing module is used to process the initial seismic signal data to form a microseismic signal database, and to add the date, time and sensor number of the seismic event to each signal data in the microseismic signal database.

[0041] The signal arrival time acquisition module is used to extract the arrival time of each signal data in the microseismic signal database using the sliding time window energy ratio method; and...

[0042] The inversion positioning module is used to locate the actual seismic source coordinates using a genetic algorithm.

[0043] In the technical solution of this invention, the initial seismic signal data is processed using the Fast Fourier Transform-Butterworth method to reduce noise and remove severely interfered signal data, successfully suppressing noise outside the seismic signal, making the seismic wave smoother, so that the initial arrival time point of the signal can be observed more clearly when the sliding time window energy ratio method is used to extract the first arrival time. Then, a genetic algorithm is used to calculate the high-precision actual seismic source coordinates, so that the seismic source positioning error is within 5%. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0045] Figure 1 A flowchart of the mine seismic source inversion method provided by the present invention;

[0046] Figure 2 for Figure 1 Flowchart of step S2;

[0047] Figure 3 for Figure 1 Flowchart of step S3;

[0048] Figure 4 for Figure 3 Flowchart of step S31;

[0049] Figure 5 for Figure 1 Flowchart of step S5;

[0050] Figure 6 for Figure 5 Flowchart of step S51;

[0051] Figure 7 for Figure 5 Flowchart of step S52;

[0052] Figure 8 This is a schematic diagram of the sliding window energy ratio method;

[0053] Figure 9 This is a schematic diagram showing the location of the actual seismic source and the sensor in a mine earthquake.

[0054] Figure 10The fast Fourier transform single-sided amplitude spectrum provided in Embodiment 1 of the present invention;

[0055] Figure 11 A schematic diagram of error analysis for Embodiment 1 provided by the present invention;

[0056] Figure 12 The fast Fourier transform single-sided amplitude spectrum provided in Embodiment 2 of the present invention;

[0057] Figure 13 A schematic diagram of error analysis for Embodiment 2 provided by the present invention;

[0058] Figure 14 The fast Fourier transform single-sided amplitude spectrum provided in Embodiment 3 of the present invention;

[0059] Figure 15 A schematic diagram of error analysis for Embodiment 3 provided by the present invention;

[0060] Figure 16 Source inversion results of the mine seismic source inversion method provided by the present invention;

[0061] Figure 17 This is a schematic diagram of an embodiment of the mine seismic source inversion system provided by the present invention.

[0062] Explanation of icon numbers:

[0063]

[0064]

[0065] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0066] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0067] It should be noted that if the embodiments of the present invention involve directional indicators (such as up, down, left, right, front, back, etc.), the directional indicators are only used to explain the relative positional relationship and movement of the components in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicators will also change accordingly.

[0068] Furthermore, if the embodiments of this invention involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the meaning of "and / or" throughout the text includes three parallel solutions; for example, "A and / or B" includes solution A, solution B, or a solution where both A and B are satisfied simultaneously. Furthermore, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

[0069] In seismic monitoring technology, source location is a key factor determining its reliability, and the accuracy of source location will have a certain impact on subsequent data interpretation. Source location refers to determining the spatial coordinates and time of the seismic source, and it is the most classic and fundamental problem in microseismic research. The basic approach to source location is based on iterative forward and inverse model inversion, using the picked first arrivals of direct P and S waves to invert the source location or time of origin, minimizing the error between the simulated and actual first arrivals. This method is currently the most widely used. In recent years, more and more new methods have been proposed, and various optimization algorithms have been applied to source location. However, current source location inversion methods still suffer from problems such as low accuracy of source inversion parameters.

[0070] In view of this, the present invention provides a method for inverting the source of a mine-induced seismic event. Figures 1 to 7 The flowchart shows the mine seismic source inversion method provided by the present invention.

[0071] Please see Figure 1 The method for inverting the seismic source of a mine earthquake includes the following steps:

[0072] Step S1: Obtain initial seismic signal data.

[0073] Further, step S1 specifically includes:

[0074] The initial seismic signal data is acquired using a coal mine microseismic monitoring system, which includes multiple sensors.

[0075] In this step, the monitoring of mine seismic signals is achieved by deploying multiple sensors in the mining area.

[0076] Step S2: Process the data using the Fast Fourier Transform-Butterworth method to obtain the waveform data of the seismic signal.

[0077] Further, please refer to Figure 2 Step S2 specifically includes:

[0078] Step S21: Normalize the initial seismic signal data to obtain normalized signal data.

[0079] In this step, the normalization process specifically involves: assuming the data received by the sensor is [x] n The normalized data is represented as [x] i Let [x] n ] mean =β, α=[x n ] max -[x n ] min Then the normalization formula is:

[0080] (1)

[0081] Normalization is the process of compressing a feature x proportionally and then shifting it. Essentially, it's a linear transformation, which doesn't change the original data's characteristics and properties; in fact, it can improve the data's performance. Normalization typically transforms a series of data into a fixed interval (range), which is a decimal between [0, 1] and (-1, 1), mainly for ease of data processing.

[0082] Step S22: Perform a fast Fourier transform on the normalized signal data to obtain the single-sided amplitude spectrum and frequency domain information of the seismic signal.

[0083] In this step, the process of performing a Fast Fourier Transform on the normalized signal data is as follows:

[0084]

[0085] In equation (2), f(t) is the initial seismic signal;

[0086] T is the period;

[0087] ω0 is the fundamental frequency, and ω0 = 2π / T;

[0088] nω0 is the harmonic frequency;

[0089] C n Let f(t) be the spectrum corresponding to the initial seismic signal f(t).

[0090] Thus, C n The output is a single-sided amplitude spectrum, which provides the frequency domain information of the seismic signal, allowing you to view the effective signal range.

[0091] Step S23: Use a Butterworth bandpass filter to filter and reduce noise in the normalized signal data to obtain the signal waveform diagram and the mine seismic signal waveform data.

[0092] In this step, since the initial seismic signal data contains a lot of noise, the sixth-order Butterworth bandpass filter with the best filtering effect is directly selected, with a sampling frequency of 500Hz.

[0093] For a Butterworth filter, the amplitude and frequency relationship of an Nth-order Butterworth low-pass filter is expressed as:

[0094]

[0095] In equation (3), H(jΩ) is the amplitude-frequency function;

[0096] Ω represents the analog angular frequency;

[0097] Ω c This is the lower boundary angular frequency of the passband;

[0098] N is the order of the Butterworth bandpass filter.

[0099] It should be noted that, as can be seen from the amplitude squared function, an Nth-order filter has 2N poles, and these 2N poles are evenly distributed on a circle, called the Butterworth circle. The Butterworth filter system is a linear system; for it to be stable, its poles must lie in the left half of the s-plane. Therefore, taking the N poles in the left half-plane as the poles of the filter ensures its stability. After determining the poles, the coefficients 'a' of the analog filter are calculated. s b s Then, by using a bilinear transformation from the analog domain to the digital domain, the coefficients a are obtained. z and b z Finally, the filtering result can be calculated using the difference equation.

[0100] More specifically, the difference equation is:

[0101]

[0102] In equation (4), M is the number of previously inputs required;

[0103] N is the number of past outputs required;

[0104] x is the input to the filter;

[0105] y is the output of the filter;

[0106] a k and b k These are the filter coefficients.

[0107] Step S3: Extract the waveform data of the mine seismic signal to form a microseismic signal database, and add the date, time and sensor number of the mine seismic event to each signal data in the microseismic signal database.

[0108] Further, please refer to Figure 3 Step S3 specifically includes:

[0109] Step S31: Extract clear actual seismic signal data from the normalized signal data, perform noise reduction processing, and integrate them to form the microseismic signal database, as shown in the table below:

[0110]

[0111] In this step, clear seismic data that can be used for experiments is obtained by extracting and denoising the normalized signal data.

[0112] For more details, please see Figure 4 Step S31 specifically includes:

[0113] Step S311: Based on the single-sided amplitude spectrum, extract clear actual mine seismic signal data from the normalized signal data and perform noise reduction processing.

[0114] In this step, the signals that are not generated by mine seismic activity are identified in the frequency domain information graph through the single-sided amplitude spectrum. For example, vibration waves generated by blasting, mining, or vehicles traveling in the mine tunnel may mask the mine seismic signal. These vibration waves are then deleted, and unclear signals in the normalized signal data are also deleted. The remaining signals are then subjected to noise reduction processing so that the final retained signal data is related to mine seismic activity.

[0115] It should be noted that the frequencies of the deleted signals are mainly concentrated around 50Hz, 100Hz, 150Hz, 200Hz and 250Hz.

[0116] Step S312: Based on the signal waveform diagram, and considering the background noise frequency range and signal-to-noise ratio, remove signal data from the actual mine seismic signal data from which noise and mine seismic waveforms cannot be separated, thereby forming the microseismic signal database.

[0117] Step S32: Add the date and time of the mine earthquake and the mine earthquake sensor number to each signal data in the microseismic signal database.

[0118] In this step, the signal data from the microseismic signal database is input into an Excel spreadsheet for organization, and then processed using Python and Pandas to generate a DataFrame for storage, which will be used in subsequent operations.

[0119] Step S4: Extract the first arrival time for each signal data in the microseismic signal database using the sliding time window energy ratio method.

[0120] This step requires controlling two parameters: the sliding window size (window_len) and the mutation threshold (Threshold). Simultaneously, the built-in Python runtime calculation toolkit (time) is used to calculate the program's execution time to improve computational speed, ensuring both high speed and maximum accuracy.

[0121] Since the data has been normalized, the data range is within (-1, 1). After multiple experiments, it was found that choosing a mutation threshold of 1.01 is more reasonable.

[0122] Step S5: Use a genetic algorithm to locate the actual seismic source coordinates.

[0123] Further, please refer to Figure 5 Step S5 specifically includes:

[0124] Step S51: Obtain the corresponding seismic source coordinates based on the position coordinates of each sensor (e.g., ...). Figure 9 (as shown), and confirmed the transmission speed of the seismic wave.

[0125] For more details, please see Figure 6 Step S51 specifically includes:

[0126] Step S511: Obtain the position coordinates of each sensor;

[0127] Step S512: Obtain the position coordinates of each sensor according to the positioning equation to obtain the corresponding seismic source position coordinates, and confirm the transmission velocity of the seismic wave. The positioning equation is:

[0128]

[0129] In equation (5), (x0, y0, z0) are the coordinates of the earthquake source location;

[0130] (x i ,y i ,z i Let be the position coordinates of the i-th sensor, where i = 1, 2, 3, ..., n;

[0131] t0 is the start time of the mine tremor;

[0132] t i This refers to the time it takes for the seismic wave to reach the sensor.

[0133] v represents the propagation speed of the seismic wave.

[0134] Furthermore, since solving the positioning equation is quite complex, it needs to be simplified accordingly, and an optimized algorithm should be used to solve it. See the following equation:

[0135]

[0136] In equation (6), t i 'The time it takes for the seismic wave to travel to the sensor;'

[0137] δ represents the average travel time residual.

[0138] Step S52: Use the genetic algorithm to invert and obtain the coordinates of the actual seismic source (e.g., Figure 9 (As shown).

[0139] Further, please refer to Figure 7 Step S52 specifically includes:

[0140] Step S521: Initialize the population.

[0141] Step S522: Initialize the coordinates of the earthquake source location.

[0142] Step S523: Calculate the fitness of each individual in the population.

[0143] Step S524: Select the individual with the highest fitness and perform crossover, mutation, and genetics.

[0144] Step S525: Update the population.

[0145] Step S526: Repeat steps S522-S525 until the maximum number of iterations is reached or a certain threshold is met to obtain the actual seismic source coordinates.

[0146] For details, please refer to Figure 1 After step S5, the method further includes:

[0147] Step S6: Perform error analysis on the actual seismic source coordinates.

[0148] In this step, an error analysis formula is used to perform error analysis on the actual seismic source coordinates. The error analysis formula is as follows:

[0149]

[0150] In equation (7), (x1, y1, z1) are the coordinate parameters calculated by the microseismic monitoring system;

[0151] (x2,y2,z2) are the actual coordinates of the seismic source in the mine.

[0152] Error analysis uses the distance error to the origin to characterize the positioning error. Since the original data uses the geodetic coordinate system, it is more meaningful to calculate the error by removing the first five digits of the x-coordinate and the first four digits of the y-coordinate.

[0153] In the technical solution of this invention, the initial seismic signal data is processed using the Fast Fourier Transform-Butterworth method to reduce noise and remove severely interfered signal data, successfully suppressing noise outside the seismic signal, making the seismic wave smoother, so that the initial arrival time point of the signal can be observed more clearly when the sliding time window energy ratio method is used to extract the first arrival time. Then, a genetic algorithm is used to calculate the high-precision actual seismic source coordinates, so that the seismic source positioning error is within 5%.

[0154] This invention provides specific embodiments to further illustrate the method for inverting seismic sources in mine earthquakes.

[0155] Example 1:

[0156] Step S1: Select the microseismic monitoring data of a high-energy mine earthquake on April 6, 2020, from a certain coal mine as the initial mine earthquake signal data.

[0157] Step S2: Process the data using the Fast Fourier Transform-Butterworth method to obtain the waveform data of the seismic signal.

[0158] Further, step S2 specifically includes:

[0159] Step S21: Compress the feature x of the initial seismic signal data proportionally, and then shift it linearly to a fixed interval (-1,1) to complete the normalized signal data processing.

[0160] In this step, normalization is used to improve the subsequent performance of the data and facilitate data processing.

[0161] Specifically, assuming the initial seismic signal data is [xn], the normalized seismic signal data is [x i Let [x] n ] mean =β, α=[x n ] max -[x n ] min Then the normalization formula is:

[0162]

[0163] Step S22: Perform a Fast Fourier Transform on the normalized signal data to obtain the single-sided amplitude spectrum of the seismic signal (e.g., Figure 10 (as shown) and frequency domain information.

[0164] In this step, for each mine tremor event, one clear signal and one unclear or relatively clear signal are selected for display. For the mine tremor event on April 6, data received by two sensors, channel 2 (unclear) and channel 20 (clear), are selected for display.

[0165] Comparative analysis revealed a significant difference between clear and unclear seismic signals. The frequencies were primarily concentrated between 1 and 30 Hz, indicating that this high-energy seismic event was a low-frequency, high-energy type. The unclear signal came from channel 2, with received frequencies mainly concentrated around 50 Hz, 100 Hz, 150 Hz, 200 Hz, and 250 Hz. These signals were not emitted by the seismic event itself; they may have been received from other sources, such as vibrations from blasting, mining, or vehicles traveling in the mine, which masked the seismic signal. The unclear signals were removed (and not used in subsequent location calculations), retaining only the relatively clear and clear signals related to the seismic event.

[0166] Step S23: Use a Butterworth bandpass filter to filter and reduce noise in the normalized signal data to obtain the signal waveform diagram and the mine seismic signal waveform data.

[0167] In this step, the upper and lower limits of the filter are selected as 1Hz and 30Hz, respectively, and signals outside this frequency range will be suppressed. This processing uses the signal.butter function from the Scipy toolkit in Python, selecting a sixth-order Butterworth bandpass filter with the best filtering effect to handle noise, with a sampling frequency of 500Hz.

[0168] Step S3: Extract the waveform data of the mine seismic signal to form a microseismic signal database, and add the date, time and sensor number of the mine seismic event to each signal data in the microseismic signal database.

[0169] Furthermore, step S3 specifically includes:

[0170] Step S31: Extract clear actual seismic signal data from the normalized signal data, perform noise reduction processing, and integrate them to form the microseismic signal database.

[0171] Further, step S31 specifically includes:

[0172] Step S311: Based on the single-sided amplitude spectrum, extract clear actual seismic signal data from the normalized signal data, including 1000 data points before and after the seismic signal, and perform noise reduction processing.

[0173] Step S312: Based on the signal waveform diagram, and considering the background noise frequency range and signal-to-noise ratio, remove signal data from the actual mine seismic signal data from which noise and mine seismic waveforms cannot be separated, to obtain 2088 sets of clear mine seismic data that can be used for testing, thus forming the microseismic signal database.

[0174] Step S32: Add the date and time of the mine earthquake and the mine earthquake sensor number to each signal data in the microseismic signal database.

[0175] This makes the study of first arrival time extraction more convenient. If comparative analysis is required, only data with the same number need to be extracted from the microseismic signal database each time.

[0176] Step S4: Extract the first arrival time for each signal data in the microseismic signal database using the sliding time window energy ratio method.

[0177] In this step, the mutation threshold Threshold is set to 1.01, the sliding window size window_len is set to 2, and the sliding window energy ratio method window is selected as follows. Figure 8 As shown.

[0178] Step S5: Use a genetic algorithm to locate the actual seismic source coordinates.

[0179] Furthermore, step S5 specifically includes:

[0180] Step S51: Obtain the corresponding seismic source coordinates based on the position coordinates of each sensor, and confirm the transmission speed of the seismic wave.

[0181] In this step, the x-coordinate range of a certain coal mine sensor is (39486000, 39494000), the y-coordinate range is (3921000, 3933000), and the z-coordinate, i.e. the depth value range, is (-800, 0) based on previous monitoring results.

[0182] Step S52: Use the genetic algorithm to invert and obtain the coordinates of the actual seismic source.

[0183] In this step, the population is first initialized. Using the range of a mine seismic sensor in a coal mine as a reference, initial values ​​for the seismic source location information are randomly assigned. The `random` function in Python is used to randomly select values ​​within the range. The initial time value is set to 0s, based on the moment the sensor records the seismic data. The initial value for the seismic wave velocity is set to 3000 m / s based on experience. The fitness of each individual in the population is calculated, and the individual with the highest fitness is selected for crossover, mutation, and genetics. The population is then updated, and the above steps are repeated until the maximum number of iterations is reached or a certain threshold is met, thus obtaining the actual seismic source coordinates (e.g., ...). Figure 16(As shown).

[0184] Step S6: Perform error analysis on the actual seismic source coordinates using error analysis formulas (e.g., ...). Figure 11 As shown in the figure, the error result obtained is 2.7%, which is within 5%.

[0185] Example 2:

[0186] Step S1: Select the microseismic monitoring data of a high-energy mine earthquake on May 15, 2020, from a certain coal mine as the initial mine earthquake signal data.

[0187] Step S2: Compress the feature x of the initial seismic signal data proportionally, and then shift it linearly to a fixed interval (-1,1) to complete the normalized signal data processing.

[0188] In this step, normalization is used to improve the subsequent performance of the data and facilitate data processing.

[0189] Specifically, assuming the initial seismic signal data is [xn], the normalized seismic signal data is [x i Let [x] n ] mean =β, α=[x n ] max -[x n ] min Then the normalization formula is:

[0190]

[0191] Further, step S2 specifically includes:

[0192] Step S21: Compress the feature x of the initial seismic signal data proportionally, and then translate it linearly to a fixed interval (-1,1) to complete the normalized signal data processing;

[0193] In this way, normalization improves the subsequent performance of the data and facilitates data processing.

[0194] Step S22: Perform a Fast Fourier Transform on the normalized signal data to obtain the single-sided amplitude spectrum of the seismic signal (e.g., Figure 12 (as shown) and frequency domain information.

[0195] In this step, for each mine tremor event, one clear signal and one unclear or relatively clear signal are selected for display. For the mine tremor event on May 15, data received by two sensors, 9 (unclear) and 15 (clear), are selected for display.

[0196] Comparative analysis revealed a significant difference between clear and unclear seismic signals. The frequencies were primarily concentrated between 1 and 30 Hz, indicating that this high-energy seismic event was a low-frequency, high-energy type. The unclear signal came from channel 9, with frequencies mainly concentrated around 50 Hz, 100 Hz, 150 Hz, 200 Hz, and 250 Hz. These signals were not emitted by the seismic event itself; they may have been received from other sources, such as vibrations from blasting, mining, or vehicles traveling in the mine, which masked the seismic signal. The unclear signals were removed (and not used in subsequent location calculations), retaining only the clearer and more relevant seismic signals.

[0197] Step S23: Use a Butterworth bandpass filter to filter and reduce noise in the normalized signal data to obtain the signal waveform diagram and the mine seismic signal waveform data.

[0198] In this step, the upper and lower limits of the filter are selected as 1Hz and 30Hz, respectively, and signals outside this frequency range will be suppressed. This processing uses the signal.butter function from the Scipy toolkit in Python, selecting a sixth-order Butterworth bandpass filter with the best filtering effect to handle noise, with a sampling frequency of 500Hz.

[0199] Step S3: Extract the waveform data of the mine seismic signal to form a microseismic signal database, and add the date, time and sensor number of the mine seismic event to each signal data in the microseismic signal database.

[0200] Furthermore, step S3 specifically includes:

[0201] Step S31: Extract clear actual seismic signal data from the normalized signal data, perform noise reduction processing, and integrate them to form the microseismic signal database.

[0202] Further, step S31 specifically includes:

[0203] Step S311: Based on the single-sided amplitude spectrum, extract clear actual seismic signal data from the normalized signal data, including 1000 data points before and after the seismic signal, and perform noise reduction processing.

[0204] Step S312: Based on the signal waveform diagram, and considering the background noise frequency range and signal-to-noise ratio, remove signal data from the actual mine seismic signal data from which noise and mine seismic waveforms cannot be separated, to obtain 2088 sets of clear mine seismic data that can be used for testing, thus forming the microseismic signal database.

[0205] Step S32: Add the date and time of the mine earthquake and the mine earthquake sensor number to each signal data in the microseismic signal database.

[0206] This makes the study of first arrival time extraction more convenient. If comparative analysis is required, only data with the same number need to be extracted from the microseismic signal database each time.

[0207] Step S4: Extract the first arrival time for each signal data in the microseismic signal database using the sliding time window energy ratio method.

[0208] In this step, the mutation threshold Threshold is set to 1.01, the sliding window size window_len is set to 2, and the sliding window energy ratio method window is selected as follows. Figure 8 As shown.

[0209] Step S5: Use a genetic algorithm to locate the actual seismic source coordinates.

[0210] Further, step S5 specifically includes:

[0211] Step S51: Obtain the corresponding seismic source coordinates based on the position coordinates of each sensor, and confirm the transmission speed of the seismic wave.

[0212] In this step, the x-coordinate range of a certain coal mine sensor is (39486000, 39494000), the y-coordinate range is (3921000, 3933000), and the z-coordinate, i.e. the depth value range, is (-800, 0) based on previous monitoring results.

[0213] Step S52: Use the genetic algorithm to invert and obtain the coordinates of the actual seismic source.

[0214] In this step, the population is first initialized. Using the range of a mine seismic sensor in a coal mine as a reference, initial values ​​for the seismic source location information are randomly assigned. The `random` function in Python is used to randomly select values ​​within the range. The initial time value is set to 0s, based on the moment the sensor records the seismic data. The initial value for the seismic wave velocity is set to 3000 m / s based on experience. The fitness of each individual in the population is calculated, and the individual with the highest fitness is selected for crossover, mutation, and genetics. The population is then updated, and the above steps are repeated until the maximum number of iterations is reached or a certain threshold is met, thus obtaining the actual seismic source coordinates (e.g., ...). Figure 16 (As shown).

[0215] Step S6: Perform error analysis on the actual seismic source coordinates using error analysis formulas (e.g., ...). Figure 13 As shown in the figure, the error result obtained is 4.8%, which is within 5%.

[0216] Example 3:

[0217] Step S1: Select the microseismic monitoring data of a large-energy mine earthquake on March 23, 2020, as the initial mine earthquake signal data.

[0218] Step S2: Compress the feature x of the initial seismic signal data proportionally, and then shift it linearly to a fixed interval (-1,1) to complete the normalized signal data processing.

[0219] In this step, normalization is used to improve the subsequent performance of the data and facilitate data processing.

[0220] Specifically, assuming the initial seismic signal data is [xn], the normalized seismic signal data is [x i Let [x] n ] mean =β, α=[x n ] max -[x n ] min Then the normalization formula is:

[0221]

[0222] Further, step S2 specifically includes:

[0223] Step S21: Compress the feature x of the initial seismic signal data proportionally, and then translate it linearly to a fixed interval (-1,1) to complete the normalized signal data processing;

[0224] In this way, normalization improves the subsequent performance of the data and facilitates data processing.

[0225] Step S22: Perform a Fast Fourier Transform on the normalized signal data to obtain the single-sided amplitude spectrum of the seismic signal (e.g., Figure 14 (as shown) and frequency domain information.

[0226] In this step, for each mine tremor event, one clear signal and one unclear or relatively clear signal are selected for display. For the mine tremor event on March 23, data received by two sensors, 0 (unclear) and 5 (clear), are selected for display.

[0227] Comparative analysis revealed a significant difference between clear and unclear seismic signals. The frequencies were primarily concentrated between 1 and 30 Hz, indicating that this high-energy seismic event was a low-frequency, high-energy type. The unclear signal was channel 0, with received frequencies mainly concentrated around 50 Hz, 100 Hz, 150 Hz, 200 Hz, and 250 Hz. These signals were not emitted by the seismic event itself; they may have been received from other sources, such as vibrations from blasting, mining, or vehicles traveling in the mine, which masked the seismic signal. The unclear signals were removed (not included in subsequent location calculations), retaining only the relatively clear and clear signals related to the seismic event.

[0228] Step S23: Use a Butterworth bandpass filter to filter and reduce noise in the normalized signal data to obtain the signal waveform diagram and the mine seismic signal waveform data.

[0229] In this step, the upper and lower limits of the filter are selected as 1Hz and 30Hz, respectively, and signals outside this frequency range will be suppressed. This processing uses the signal.butter function from the Scipy toolkit in Python, selecting a sixth-order Butterworth bandpass filter with the best filtering effect to handle noise, with a sampling frequency of 500Hz.

[0230] Step S3: Extract the waveform data of the mine seismic signal to form a microseismic signal database, and add the date, time and sensor number of the mine seismic event to each signal data in the microseismic signal database.

[0231] Furthermore, step S3 specifically includes:

[0232] Step S31: Extract clear actual seismic signal data from the normalized signal data, perform noise reduction processing, and integrate them to form the microseismic signal database.

[0233] Further, step S31 specifically includes:

[0234] Step S311: Based on the single-sided amplitude spectrum, extract clear actual seismic signal data from the normalized signal data, including 1000 data points before and after the seismic signal, and perform noise reduction processing.

[0235] Step S312: Based on the signal waveform diagram, and considering the background noise frequency range and signal-to-noise ratio, remove signal data from the actual mine seismic signal data from which noise and mine seismic waveforms cannot be separated, to obtain 2088 sets of clear mine seismic data that can be used for testing, thus forming the microseismic signal database.

[0236] Step S32: Add the date and time of the mine earthquake and the mine earthquake sensor number to each signal data in the microseismic signal database.

[0237] This makes the study of first arrival time extraction more convenient. If comparative analysis is required, only data with the same number need to be extracted from the microseismic signal database each time.

[0238] Step S4: Extract the first arrival time for each signal data in the microseismic signal database using the sliding time window energy ratio method.

[0239] In this step, the mutation threshold Threshold is set to 1.01, the sliding window size window_len is set to 2, and the sliding window energy ratio method window is selected as follows. Figure 8 As shown.

[0240] Step S5: Use a genetic algorithm to locate the actual seismic source coordinates.

[0241] Furthermore, step S5 specifically includes:

[0242] Step S51: Obtain the corresponding seismic source coordinates based on the position coordinates of each sensor, and confirm the transmission speed of the seismic wave.

[0243] In this step, the x-coordinate range of a certain coal mine sensor is (39486000, 39494000), the y-coordinate range is (3921000, 3933000), and the z-coordinate, i.e. the depth value range, is (-800, 0) based on previous monitoring results.

[0244] Step S52: Use the genetic algorithm to invert and obtain the coordinates of the actual seismic source.

[0245] In this step, the population is first initialized. Using the range of a mine seismic sensor in a coal mine as a reference, initial values ​​for the seismic source location information are randomly assigned. The `random` function in Python is used to randomly select values ​​within the range. The initial time value is set to 0s, based on the moment the sensor records the seismic data. The initial value for the seismic wave velocity is set to 3000 m / s based on experience. The fitness of each individual in the population is calculated, and the individual with the highest fitness is selected for crossover, mutation, and genetics. The population is then updated, and the above steps are repeated until the maximum number of iterations is reached or a certain threshold is met, thus obtaining the actual seismic source coordinates (e.g., ...). Figure 16 (As shown).

[0246] Step S6: Perform error analysis on the actual seismic source coordinates using error analysis formulas (e.g., ...). Figure 15 As shown in the figure, the error result obtained is 4.4%, which is within 5%.

[0247] This invention also provides a mine-induced seismic source inversion system 100, applicable to the aforementioned mine-induced seismic source inversion method. Please refer to [link / reference needed]. Figure 17 The seismic source inversion system 100 includes a seismic signal acquisition module 1, a signal processing module 2, a signal arrival time acquisition module 3, and an inversion positioning module 4. The seismic signal acquisition module 1 is used to acquire initial seismic signal data; the signal processing module 2 is used to process the initial seismic signal data to form a microseismic signal database, and to add the seismic occurrence date, time, and seismic sensor number to each signal data in the microseismic signal database; the signal arrival time acquisition module 3 is used to extract the arrival time of each signal data in the microseismic signal database using the sliding window energy ratio method; and the inversion positioning module 4 is used to locate the actual seismic source coordinates using a genetic algorithm.

[0248] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural transformations made using the contents of the present invention's specification and drawings under the inventive concept of the present invention, or direct / indirect applications in other related technical fields, are included within the patent protection scope of the present invention.

Claims

1. A method for inverting the source of a mine-induced seismic event, characterized in that, The method for inverting the seismic source of a mine earthquake includes the following steps: Step S1: Obtain initial seismic signal data; Step S2: Process the data using the Fast Fourier Transform-Butterworth method to obtain the waveform data of the mine seismic signal; Step S3: Extract the waveform data of the seismic signal to form a microseismic signal database, and add the date, time and sensor number of the seismic event to each signal data in the microseismic signal database; Step S4: Extract the first arrival time for each signal data in the microseismic signal database using the sliding window energy ratio method; Step S5: Use a genetic algorithm to locate the actual seismic source coordinates; Step S2 specifically includes: Step S21: Normalize the initial seismic signal data to obtain normalized signal data; Step S22: Perform a fast Fourier transform on the normalized signal data to obtain the single-sided amplitude spectrum and frequency domain information of the seismic signal; Step S23: Use a Butterworth bandpass filter to filter and reduce noise in the normalized signal data to obtain the signal waveform diagram and the mine vibration signal waveform data; Following step S5, the following is also included: Step S6: Perform error analysis on the actual seismic source coordinates.

2. The method for inverting the seismic source of a mine earthquake as described in claim 1, characterized in that, Step S3 specifically includes: Step S31: Extract clear actual seismic signal data from the normalized signal data, perform noise reduction processing, and integrate them to form the microseismic signal database; Step S32: Add the date and time of the mine earthquake and the mine earthquake sensor number to each signal data in the microseismic signal database.

3. The method for inverting the seismic source of a mine earthquake as described in claim 2, characterized in that, Step S31 specifically includes: Step S311: Based on the single-sided amplitude spectrum, extract clear actual mine seismic signal data from the normalized signal data and perform noise reduction processing; Step S312: Based on the signal waveform diagram, and considering the background noise frequency range and signal-to-noise ratio, remove signal data from the actual mine seismic signal data from which noise and mine seismic waveforms cannot be separated, thereby forming the microseismic signal database.

4. The method for inverting the seismic source of a mine earthquake as described in claim 1, characterized in that, Step S1 specifically includes: The initial seismic signal data is acquired using a coal mine microseismic monitoring system, wherein the coal mine microseismic monitoring system includes multiple sensors; Step S5 specifically includes: Step S51: Obtain the corresponding seismic source coordinates based on the position coordinates of each sensor, and confirm the transmission speed of the seismic wave. Step S52: Use the genetic algorithm to invert and obtain the coordinates of the actual seismic source.

5. The method for inverting the seismic source of a mine earthquake as described in claim 4, characterized in that, Step S51 specifically includes: Step S511: Obtain the position coordinates of each sensor; Step S512: Obtain the position coordinates of each sensor according to the positioning equation to obtain the corresponding seismic source position coordinates, and confirm the transmission velocity of the seismic wave. The positioning equation is as follows: Where (x0, y0, z0) are the coordinates of the earthquake source location, (x i ,y i ,z i Let t0 be the position coordinates of the i-th sensor, i = 1, 2, 3, ..., n, and t0 be the start time of the mine tremor. i denoted as , where is the time it takes for the seismic wave to reach the sensor; v is the transmission speed of the seismic wave.

6. The method for inverting the seismic source of a mine earthquake as described in claim 4, characterized in that, Step S52 specifically includes: Step S521: Initialize the population; Step S522: Initialize the coordinates of the earthquake source location; Step S523: Fitness of each individual in the population; Step S524: Select the individual with the highest fitness and perform crossover, mutation, and genetics. Step S525: Update the population; Step S526: Repeat steps S522-S525 until the maximum number of iterations is reached or a certain threshold is met to obtain the actual seismic source coordinates.

7. A mine-induced seismic source inversion system, applicable to the mine-induced seismic source inversion method as described in any one of claims 1-6, characterized in that, The seismic source inversion system for mine earthquakes includes: The seismic signal acquisition module is used to acquire initial seismic signal data; The signal processing module is used to process the initial seismic signal data to form a microseismic signal database, and to add the seismic occurrence date, time and seismic sensor number to each signal data in the microseismic signal database; The signal arrival time acquisition module is used to extract the arrival time of each signal data in the microseismic signal database using the sliding time window energy ratio method; and... The inversion positioning module is used to locate the actual seismic source coordinates using a genetic algorithm.

Images