A method for early warning of rock burst based on active and passive dual-source combined shock wave CT inversion

Abstract

The present invention discloses a method for inversion of shockwave CT and rockburst early warning based on active and passive dual-source combined shockwave CT. The method comprises the following steps: 1. constructing a probabilistic non-uniform initial velocity distribution model of longitudinal and / or shear waves and an initial mesh model of a working face; 2. performing forward traveltime tomography to calculate the theoretical traveltime and residual of the theoretical longitudinal and / or shear waves; 3. solving the gradient of the objective function of the traveltime residual; 4. performing traveltime tomography inversion to iteratively adjust the velocity parameters of the source and longitudinal and / or shear waves of the shockwaves, so that the residual and RMS residual reach a stable state, and outputting the final source parameters and wave velocity; 5. testing the feasibility and accuracy of the inversion model through active CT numerical simulation, passive CT numerical simulation, and active and passive dual-source combined CT numerical simulation experiments. This method makes the inverted working face wave velocity distribution more accurate, thereby improving the efficiency and accuracy of the shockwave CT technology.

Description

Technical Field

[0001] The present invention relates to an active-passive dual-source combined shock wave CT inversion method and a rock burst early warning method. Background Art

[0002] Vibration wave CT technology is a key technology in the field of mine safety monitoring. By analyzing the changes in the propagation velocity of vibration waves in coal and rock masses and inverting the distribution of underground stress fields, this technology can more accurately reflect the regional stress field, thereby predicting the occurrence of dynamic disasters such as rock bursts. This provides a scientific basis for the prediction of microseismic events in space and in the future, and is of great significance for guiding mine safety production. Although vibration wave CT technology has demonstrated certain advantages in detecting rock burst risks in coal mines, its application also faces several challenges:

[0003] First of all, with the continuous deepening of coal resource mining, deep mining has become an important trend in the coal industry. However, with the increase in mining depth, the ground stress on the coal mine rock mass also increases significantly. The working conditions under the coal mine underground are more complex and are greatly affected by the working environment. In addition, the coal seam contains many factors that affect the propagation of seismic waves. Therefore, when performing CT scanning on the coal rock layer, it is often impossible to scan all rock layers, and different inversion models often need to be established during inversion.

[0004] Secondly, the shortcomings of coal and rock CT data volume, slow processing speed, etc. are also important factors restricting the application of vibration wave CT technology in coal mine rock burst risk detection.

[0005] In summary, although shock wave CT technology has good detection effects, it also has certain limitations. In particular, different inversion models need to be established during inversion. In addition, the large amount of data and slow processing speed have seriously restricted the application of shock wave CT technology in rock burst warning. Summary of the Invention

[0006] In response to the above problems, the present invention provides an active and passive dual-source combined shock wave CT inversion method and an impact ground pressure warning method, develops an inversion model and algorithm suitable for the active and passive dual-source combined CT technology, realizes the correction of the single shock wave CT imaging result, makes the inverted working face wave velocity distribution more accurate, improves the efficiency and accuracy of the shock wave CT technology, and provides scientific guidance for pressure relief measures such as large-diameter drilling and coal rock blasting.

[0007] In order to achieve the above technical objectives and the above technical effects, the present invention is implemented through the following technical solutions:

[0008] A method for inversion of shock wave CT based on active and passive dual-source combined methods includes the following steps:

[0009] Step 1: construct a probabilistic non-uniform initial velocity distribution model of longitudinal and / or shear waves of the seismic wave according to the structural characteristics of the working face, establish an initial grid model of the working face, and set the positions of the active and passive seismic sources and the positions of the seismic sensors for detecting the active and passive seismic waves;

[0010] Step 2: Perform travel time tomography forward modeling using the initial models obtained in step 1 to calculate theoretical travel times and residuals of theoretical longitudinal waves and / or shear waves;

[0011] Step 3: Based on the theoretical travel time and residual value obtained in step 2, the gradient of the objective function of the shock wave travel time residual is solved;

[0012] Step 4: Performing seismic wave travel time tomography inversion on the gradient of the objective function obtained in step 3, iteratively adjusting the velocity parameters of the source and the longitudinal and / or shear waves of the seismic wave until the residual and the RMS residual reach a stable state, and outputting the final source parameters and wave velocity;

[0013] Step 5: Verify the feasibility and accuracy of the inversion model through active CT numerical simulation, passive CT numerical simulation, and active-passive dual-source combined CT numerical simulation experiments: When the active CT numerical simulation, passive CT numerical simulation, and active-passive dual-source combined CT numerical simulation all meet the feasibility and accuracy, the inversion ends.

[0014] Preferably, in step 1, the step of constructing a probabilistic non-uniform initial velocity distribution model includes:

[0015] Step 101: Establish an initial coal mine tunnel model that simulates the actual conditions of the coal mine tunnel at the working face, and preliminarily calculate the observed wave velocity of each longitudinal wave and / or shear wave by dividing the distance by the duration based on the distance between the active and passive seismic sources and the seismic pickup sensor and the first arrival travel time of the seismic wave;

[0016] Step 102: Using a random number algorithm, a value within the observed wave velocity range is selected as a probabilistic non-uniform initial velocity and assigned to the initial shock wave travel time tomography inversion model;

[0017] Step 103: Perform inversion numerical simulation to verify the accuracy of the initial velocity distribution model through active CT numerical simulation, passive CT numerical simulation, and active-passive dual-source combined CT numerical simulation.

[0018] Step 104: When the active CT numerical simulation, the passive CT numerical simulation, and the active-passive dual-source combined CT numerical simulation all meet the accuracy requirements, the inversion ends, and the obtained initial velocity distribution model is a probabilistic non-uniform initial velocity distribution model of the longitudinal and / or shear waves of the vibration wave.

[0019] Preferably, in step 3, the specific calculation steps include:

[0020] Step 301: Define the shock wave travel time residual E as the objective function of the travel time difference:

[0021]

[0022] Where E(c, t0, s) is the residual of the shock wave travel time; c is the velocity of the longitudinal and / or shear wave propagation of the shock wave, t0 is the starting time of the earthquake source, and s is the distance the shock wave propagates; T(c, g; t0, s x , s y , s z ) is the earthquake source location (s x , s y , s z ) is the theoretical time it takes for the vibration wave to propagate to the vibration sensor g; T(c, g; t0, s x , s y , s z )=T(c,g;s x , s y , s z )+t0,(s x , s y , s z ) is the coordinate of the earthquake source position, T * (g) is the observed propagation time of the seismic pickup sensor g relative to the chosen zero time;

[0023] Step 302: If the velocity c of the longitudinal and / or shear wave propagation and the source parameter (s x , s y , s z , t0) changes cause the calculated arrival time T of the transmitting signal source to change accordingly, so the change of the travel time residual is for:

[0024]

[0025] Step 303: The partial derivatives of the residual E with respect to the position, origin time and speed change are:

[0026]

[0027] Among them, s i =s x , s y or s z , λ is the Lagrange multiplier;

[0028] Step 304: Passing at the border Initialize λ with the boundary value and set the remaining grid points to zero;

[0029] Step 305: Alternately scan the entire domain using Update the λ value of non-boundary positions;

[0030] Step 306: For a given convergence criterion ε>0, check the criterion ||λ point by point. n+1 -λ n || L1 ≤ε to check convergence;

[0031] Step 307: When the convergence is satisfied, the corresponding The value is used as the gradient of the objective function;

[0032] in, is the normal vector perpendicular to the surface of the vibration pickup sensor, is the gradient, T is the calculated arrival time of the transmitting signal source, λ is the Lagrange multiplier, T° refers to the propagation time of the longitudinal and / or shear waves of the vibration waves observed by the vibration pickup sensor, Ω is the vibration pickup sensor surface, x∈Ω refers to the position x on the vibration pickup sensor surface, ||λ n+1 -λ n || L1 is the Lagrange multiplier matrix L 1 norm, indicating that the algorithm converges when the rate of change of the Lagrange multiplier is negative, λ n+1 is the λ value obtained in the n+1th iteration, λ n is the value of λ obtained at the nth iteration.

[0033] Preferably, in step 4, the step of outputting the final source parameters and wave velocity is:

[0034] When the shock wave travels, the residual E and RMS residual If a stable state is reached, the inversion is judged to be converged, and the final source parameters and wave velocity are output, where N is the number of observed vibrations; otherwise, step 302 is entered for iteration.

[0035] Preferably, the probabilistic non-uniform initial velocity ranges from 1000 m / s to 6500 m / s.

[0036] Correspondingly, a rock burst warning method uses the inversion model obtained by any of the above-mentioned active and passive dual-source combined shock wave CT inversion methods to perform rock burst warning.

[0037] The beneficial effects of the present invention are:

[0038] First, the active-passive dual-source combined CT inversion method of the present invention improves the limitations of a single imaging method and adopts an adjoint state algorithm based on rapid scanning to improve the convergence speed of the inversion model and the accuracy of the inversion results, providing a powerful reference for the prediction of coal mine rock burst hazards.

[0039] Second, the probabilistic non-uniform initial velocity distribution model of the present invention has high accuracy and strong practicality, and has an excellent effect on the optimization of coal mine vibration wave tomography technology. The velocity distribution obtained by inversion is closer to the actual situation, and can better predict the high-stress dangerous areas of the mining working face, providing a basis for the prediction and prevention of rock burst risks.

[0040] Third, the present invention directly obtains the gradient of the objective function and adopts the gradient method and nonlinear optimization algorithm to avoid the traditional direct solution of the Fréchet equation, which avoids the shortcomings of its slow operation speed and unsuitability for large-scale model data processing.

[0041] Fourth, the present invention integrates the two major advantages of active CT in detection accuracy and passive CT in detection timeliness, and eliminates to the maximum extent the shortcomings and disadvantages of active CT inversion technology in detection timeliness and its detection process interference with production, as well as the deficiencies and shortcomings of passive CT inversion technology in inversion accuracy, thereby realizing real-time detection and inversion of fast short-interval continuous wave velocity. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 This is an overall flow chart of the present invention's method for inverting active and passive dual-source combined shock wave CT;

[0043] Figure 2 This is a rendering of the coal mine tunnel model of the present invention;

[0044] Figure 3 2. It is a schematic diagram comparing the effects of the uniform initial velocity distribution model and the probabilistic non-uniform initial velocity distribution model of the present invention;

[0045] Figure 4 2. It is a schematic diagram comparing the effects of the shock wave tomography simulation of the present invention;

[0046] Figure 5 This is a layout diagram of artificial blasting points and vibration pickup sensors on a working face according to an embodiment of the present invention;

[0047] Figure 6 This is a distribution diagram of probabilistic non-uniform initial velocity values ​​of a working face according to an embodiment of the present invention;

[0048] Figure 7 This is a wave velocity contour map of the working face vibration wave tomography according to an embodiment of the present invention;

[0049] Figure 8 This is the model and effect diagram of the active CT inversion numerical simulation of the present invention;

[0050] Figure 9 This is the model and effect diagram of the passive CT inversion numerical simulation of the present invention;

[0051] Figure 10This is the model and effect diagram of the active and passive dual-source combined CT inversion numerical simulation of the present invention. DETAILED DESCRIPTION

[0052] The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments so that those skilled in the art can better understand the present invention and implement it, but the embodiments are not intended to limit the present invention.

[0053] like Figure 1 As shown, a method for inversion of active and passive dual-source combined shock wave CT includes the following steps:

[0054] Step 1: Construct a probabilistic non-uniform initial velocity distribution model of longitudinal and / or transverse vibration waves based on the structural characteristics of the working face, establish an initial grid model of the working face, and set the positions of the active and passive seismic sources and the positions of the vibration pickup sensors for detecting the active and passive vibration waves. Among them, the artificial seismic source corresponds to the active seismic source, and the natural seismic source corresponds to the passive seismic source.

[0055] As an important geophysical method, shock wave tomography analyzes the differences in shock wave propagation velocities in different media and measures the propagation time of shock waves through coal and rock masses to invert the velocity structure within the coal and rock masses. This method has excellent applicability and accuracy for assessing rock burst risk. This method can reveal stress concentration areas and high-risk regions within the coal and rock masses, providing a scientific basis for rock burst prediction and prevention. Research has shown that shock wave velocity is closely related to the stress state of the coal and rock masses, with high wave velocity areas often corresponding to high stress areas and potential areas for rock burst.

[0056] In the existing technology, there is relatively little research on optimizing the initial velocity model for shock wave tomography. Most inversion studies use the mean velocity model, which assumes that the initial velocity remains constant. However, the velocity of shock waves varies when passing through different rock masses, and a simple mean velocity model cannot simulate the actual geological environment of a coal mine working face. Therefore, to improve the accuracy of shock wave tomography technology, it is necessary to further optimize the initial velocity model to reflect the changes in the velocity of shock waves passing through different coal and rock masses, so as to adapt to the complex geological and production environment of coal mines and improve the accuracy of rock burst risk prediction.

[0057] The principle of CT inversion is to invert and reconstruct the velocity model using distance and received first arrival times. The initial velocity model is unknown, and its setting is particularly important for inversion accuracy. The closer the initial velocity model is to the actual velocity model, the more accurate the inversion will be. To this end, the present invention proposes a probabilistic non-uniform initial velocity distribution model for seismic waves, longitudinal waves, and / or shear waves, constructed based on the structural characteristics of the working face strata. The construction steps are as follows:

[0058] Step 101: Establish an initial coal mine tunnel model that simulates the actual situation of the coal mine tunnel on the working face. According to the distance between the active and passive seismic sources and the seismic sensors and the initial arrival time of the vibration wave, preliminarily calculate the observed wave velocity of each vibration wave longitudinal wave (P wave) and / or shear wave (S wave) by dividing the distance by the duration.

[0059] The purpose of using coal mine tunnel models is to better reflect the actual situation of the mine working face, such as Figure 2 As shown, a low-velocity region in the shape of an elliptical cylinder in the XZ direction is set in the middle of the seismic sensor deployment range. The radius of the elliptical cylinder's minor axis is 20m, the radius of the major axis is 30m, and the height of the cylinder is 100m. The velocity within this cylinder is set to 4500m / s as the low-velocity region to simulate the velocity anomaly, while the velocity of the remaining area is set to 5500m / s as the high-velocity region. This serves as the initial velocity model to simulate the conditions inside the tunnel, and a control experiment is then conducted. Based on the distance between the artificial or natural seismic source and the seismic sensor, as well as the initial arrival travel time of the seismic wave, the observed velocity of each seismic wave is preliminarily calculated by dividing the distance by the duration.

[0060] Step 102: Using a random number algorithm, a value within the observed wave velocity range is selected as a probabilistic non-uniform initial velocity and assigned to an initial shock wave travel time tomography inversion model.

[0061] It is known that the acoustic wave velocity of coal seams is between 1990 and 3000 m / s, and the acoustic wave velocity of surrounding rocks is between 2300 and 5100 m / s. In order to better reduce the error caused by the vibration wave received by the station, the range of randomly extracted wave velocity is set to 1000 m / s-6500 m / s. After screening, the wave velocity that is too high or too low due to some reasons can be removed. Under real geological conditions, the extracted wave velocity is generally in line with the normal distribution. The generated mean velocity model (i.e., equal value uniform distribution model) and probabilistic non-uniform velocity model are as follows: Figure 3 shown.

[0062] Step 103: Perform inversion numerical simulation on the inversion model obtained in step 102, and verify the accuracy of the initial velocity distribution model through active CT numerical simulation, passive CT numerical simulation, and active-passive dual-source combined CT numerical simulation.

[0063] Step 104: When the active CT numerical simulation, the passive CT numerical simulation, and the active-passive dual-source combined CT numerical simulation all meet the accuracy requirements, the inversion ends, and the obtained initial velocity distribution model is a probabilistic non-uniform initial velocity distribution model of the longitudinal and / or shear waves of the vibration wave.

[0064] The simulation results of the uniform velocity distribution model in the three tomography methods in step 103 are compared to verify the superiority of the probabilistic non-uniform velocity model in terms of restoration degree, such as Figure 4 As shown, where:

[0065] In the active CT numerical simulation results, the residual of the uniform velocity distribution model decreased from 213.104 to 0.322, and the RMS decreased from 0.703 to 0.027. However, the residual of the non-uniform initial velocity model decreased from 195.716 to 0.239, and the RMS decreased from 0.670 to 0.023. With the same number of iterations, the non-uniform initial velocity model had smaller values ​​than the mean initial velocity model, both in terms of the initial residual value and the convergence effect of the final iteration, indicating a better performance.

[0066] In the passive CT numerical simulation results, the residual of the uniform velocity distribution model decreased from 253.445 to 6.667, and the RMS decreased from 0.3190 to 0.0517. However, the residual of the non-uniform initial velocity model decreased from 190.631 to 1.831, and the RMS decreased from 0.2766 to 0.0271. With the same number of iterations, the non-uniform initial velocity model had smaller values ​​than the mean initial velocity model, both in terms of the initial residual value and the convergence effect of the final iteration, indicating a better performance.

[0067] In the numerical simulation results of active and passive dual-source CT, the residual of the uniform velocity distribution model decreased from 167.995 to 3.297, and the RMS decreased from 0.2597 to 0.0364; while the residual of the non-uniform initial velocity model decreased from 119.736 to 2.3551, and the RMS decreased from 0.2023 to 0.0283. The residual iteration effect of the non-uniform initial velocity model is better than that of the mean initial velocity model.

[0068] The example application selects a coal mine working face as the research object, such as Figure 5 The following figure shows the layout of the tunnels at the working face. Figure 5 The figure shows the locations of artificial shot points and seismic sensors (i.e., geophones). Shot points were placed on the lower wall of the working face, with a spacing of 18 meters between them, for a total of 33 shot points. Geophones were placed on the upper wall of the lower roadway, with a spacing of 9 meters (15 times the bolt spacing), for a total of 66 receiving points. The initial velocity model of the studied working face was established using the initial velocity distribution optimization model for seismic wave tomography presented in the present invention. Active CT, passive CT, and active-passive dual-source CT were used to invert and analyze the velocity anomaly areas of the working face. Finally, the accuracy of the inversion results was verified using the distribution of microseismic sources during mining.

[0069] The initial wave velocity model used in the inversion is based on the probability non-uniform initial velocity distribution. Most of the shock waves generated by the shot point pass through the single medium of the coal seam and are received by the sensor. The non-uniform initial wave velocity range is set to 1000m / s-6500m / s. The wave velocity distribution is as follows: Figure 6 shown.

[0070] like Figure 7 As shown, according to the initial velocity model, the adjoint state method based on FSM (fast scanning method) is used to perform active, passive, and active-passive dual-source combined CT inversion on the model of the embodiment. The high wave velocity area above 3800m / s has been Figure 7 The results show that the earthquake sources induced during the mining process are basically distributed in the high wave velocity area and the wave velocity change area, and the active and passive dual-source combined CT inversion can correct the passive CT inversion wave velocity image, which is more consistent with the distribution of mining earthquake sources.

[0071] The initial velocity distribution optimization model for shock wave tomography, described in this invention, overcomes the limitations of the mean distribution model used in inversion for initial velocity distribution, making the velocity distribution more consistent with the actual mine geological environment. The inversion results can be accurately applied to various shock wave tomography schemes for coal mine roadway models. This method can make the inverted velocity distribution of the working face more accurate and more consistent with the actual working face conditions, providing scientific guidance for implementing pressure relief measures in rock burst mines and possessing considerable practical significance.

[0072] Step 2: Perform travel time tomography forward modeling using the initial models obtained in step 1 to calculate theoretical travel times and residuals of theoretical longitudinal waves and / or shear waves;

[0073] Step 3: Based on the theoretical travel time and residual value obtained in step 2, the gradient of the objective function of the shock wave travel time residual is solved;

[0074] Step 4: Performing seismic wave travel time tomography inversion on the gradient of the objective function obtained in step 3, iteratively adjusting the velocity parameters of the source and the longitudinal and / or shear waves of the seismic wave until the residual and the RMS residual reach a stable state, and outputting the final source parameters and wave velocity;

[0075] Step 5: Verify the feasibility and accuracy of the inversion model through active CT numerical simulation, passive CT numerical simulation, and active-passive dual-source combined CT numerical simulation experiments: When the active CT numerical simulation, passive CT numerical simulation, and active-passive dual-source combined CT numerical simulation all meet the feasibility and accuracy, the inversion ends.

[0076] In step 1: establish an initial grid model of the working face, and set the positions of the active and passive seismic sources and the positions of the seismic sensors that detect the active and passive vibration waves: for example, establish a 3D rectangular model simulating underground rock conditions, with a size of 60m×10m×100m, in which 30 seismic sensors with good azimuth coverage are arranged.

[0077] Preferably, in step 3, the specific calculation steps include:

[0078] Step 301: Define the shock wave travel time residual E as the objective function of the travel time difference:

[0079]

[0080] Where E(c, t0, s) is the residual of the shock wave travel time; c is the velocity of the longitudinal and / or shear wave propagation of the shock wave, t0 is the starting time of the earthquake source, and s is the distance the shock wave propagates; T(c, g; t0, s x , s y , s z ) is the earthquake source location (s x , s y , s z ) is the theoretical time it takes for the vibration wave to propagate to the vibration sensor g; T(c, g; t0, s x , s y , s z )=T(c,g;s x , s y , s z )+t0,(s x , s y , s z ) is the coordinate of the earthquake source position, T * (g) is the observed propagation time of the seismic pickup sensor g relative to the chosen zero time;

[0081] Step 302: If the velocity c of the longitudinal and / or shear wave propagation and the source parameter (s x , s y , s z , t0) causes the calculated arrival time T of the transmitting signal source to change accordingly (if it does not cause a change, the residual value is 0), then the change in the travel time residual is for:

[0082]

[0083] Step 303: The partial derivatives of the residual E with respect to the position, origin time and speed change are:

[0084]

[0085] Among them, s i =s x , s y or s z , λ is the Lagrange multiplier;

[0086] Step 304: Passing at the border Initialize λ with the boundary value and set the remaining grid points to zero;

[0087] Step 305: Alternately scan the entire domain using Update the λ value of non-boundary positions;

[0088] Step 306: For a given convergence criterion ε>0, check the criterion ||λ point by point. n+1 -λ n || L1 ≤ε to check convergence;

[0089] Step 307: When the convergence is satisfied, the corresponding The value is used as the gradient of the objective function;

[0090] in, is the normal vector perpendicular to the surface of the vibration pickup sensor, is the gradient, T is the calculated arrival time of the transmitting signal source, λ is the Lagrange multiplier, T ° It refers to the propagation time of the longitudinal and / or shear waves of the vibration waves observed by the vibration pickup sensor, Ω is the surface of the vibration pickup sensor, x∈Ω refers to the position x on the surface of the vibration pickup sensor, ‖λ nt1 -λ n ‖ L1 is the Lagrange multiplier matrix L 1 norm, indicating that the algorithm converges when the rate of change of the Lagrange multiplier is negative, λ n+1 is the λ value obtained in the n+1th iteration, λ n is the value of λ obtained at the nth iteration.

[0091] Preferably, in step 4, considering that the residual E has different degrees of sensitivity to changes in t0, s and c, at the end of the iteration, when the shock wave travels, the residual E and the RMS residual If a stable state is reached, that is, the percentage of reduction is less than the target threshold, the inversion is judged to be converged, and the final source parameters and wave velocity are output, where N is the number of observed vibrations; otherwise, step 302 is entered for iteration.

[0092] Through active CT (such as Figure 8 As shown), passive CT (as Figure 9 As shown), active and passive dual-source combined CT (as shown Figure 10 Numerical simulations (shown in Figure 2) verified the feasibility and accuracy of the inversion model, demonstrating that the regional distribution of velocity anomalies matched the true source location and that the residuals converged to ideal values. In the numerical simulation of active and passive dual-source CT inversion, the mutual ray reception model used in active CT inversion was changed to one in which the station locations were known sources. A randomly perturbed passive source was added, imposing random perturbations whose time and position followed a standard normal distribution to simulate the randomness of in-situ microseismic events.

[0093] Examples demonstrate that the inversion method of the present invention is highly feasible, simple to acquire data, provides timely and clear analysis results, and is highly practical. It can effectively predict high-stress hazardous areas within a mining face, providing a reference for implementing pressure relief measures. Furthermore, verification has shown that almost all earthquake sources are located in high-velocity and velocity-varying regions. Therefore, the inversion model can be applied to rock burst warning. Accordingly, a rock burst warning method employs any of the above-described inversion models obtained using the active-passive dual-source combined seismic wave CT inversion methods for rock burst warning.

[0094] The above are only preferred embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structures or equivalent process changes made using the contents of the present invention's description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for inversion of shock wave CT based on active and passive dual-source combined, characterized in that: The steps include: Step 1: construct a probabilistic non-uniform initial velocity distribution model of longitudinal and / or shear waves of the seismic wave according to the structural characteristics of the working face, establish an initial grid model of the working face, and set the positions of the active and passive seismic sources and the positions of the seismic sensors for detecting the active and passive seismic waves; Step 2: Perform travel time tomography forward modeling using the initial models obtained in step 1 to calculate theoretical travel times and residuals of theoretical longitudinal waves and / or shear waves; Step 3: Based on the theoretical travel time and residual value obtained in step 2, the gradient of the objective function of the shock wave travel time residual is solved; Step 4: Performing seismic wave travel time tomography inversion on the gradient of the objective function obtained in step 3, iteratively adjusting the velocity parameters of the source and the longitudinal and / or shear waves of the seismic wave until the residual and the RMS residual reach a stable state, and outputting the final source parameters and wave velocity; Step 5: Verify the feasibility and accuracy of the inversion model through active CT numerical simulation, passive CT numerical simulation, and active-passive dual-source combined CT numerical simulation experiments: When the active CT numerical simulation, passive CT numerical simulation, and active-passive dual-source combined CT numerical simulation all meet the feasibility and accuracy, the inversion ends.

2. The method for inversion of shock wave CT based on active and passive dual-source combined according to claim 1 is characterized in that: In step 1, the steps of constructing the probabilistic non-uniform initial velocity distribution model include: Step 101: Establish an initial coal mine tunnel model that simulates the actual conditions of the coal mine tunnel at the working face, and preliminarily calculate the observed wave velocity of each longitudinal wave and / or shear wave by dividing the distance by the duration based on the distance between the active and passive seismic sources and the seismic pickup sensor and the first arrival travel time of the seismic wave; Step 102: Using a random number algorithm, a value within the observed wave velocity range is selected as a probabilistic non-uniform initial velocity and assigned to the initial shock wave travel time tomography inversion model; Step 103: Perform inversion numerical simulation to verify the accuracy of the initial velocity distribution model through active CT numerical simulation, passive CT numerical simulation, and active-passive dual-source combined CT numerical simulation. Step 104: When the active CT numerical simulation, the passive CT numerical simulation, and the active-passive dual-source combined CT numerical simulation all meet the accuracy requirements, the inversion ends, and the obtained initial velocity distribution model is a probabilistic non-uniform initial velocity distribution model of the longitudinal and / or shear waves of the vibration wave.

3. The method for inversion based on active and passive dual-source combined shock wave CT according to claim 2 is characterized in that: In step 3, the specific calculation steps include: Step 301: Define the shock wave travel time residual E as the objective function of the travel time difference: Where E(c, t0, s) is the residual of the shock wave travel time; c is the velocity of the longitudinal and / or shear wave propagation of the shock wave, t0 is the starting time of the earthquake source, and s is the distance the shock wave propagates; T(c, g; t0, s x , s y , s z ) is the earthquake source location (s x , s y , s z ) is the theoretical time it takes for the vibration wave to propagate to the vibration sensor g; T(c, g; t0, s x , s y , s z )=T(c,g;s x , s y , s z )+t0,(s x , s y , s z ) is the coordinate of the earthquake source position, T * (g) is the observed propagation time of the seismic pickup sensor g relative to the chosen zero time; Step 302: If the velocity c of the longitudinal and / or shear wave propagation and the source parameter (s x , s y , s z , t0) changes cause the calculated arrival time T of the transmitting signal source to change accordingly, so the change of the travel time residual is for: Step 303: The partial derivatives of the residual E with respect to the position, origin time and speed change are: Among them, s i =s x , s y or s z , λ is the Lagrange multiplier; Step 304: Passing at the border (x∈Ω), initialize λ with the boundary value and set the remaining grid points to zero; Step 305: Alternately scan the entire domain using (x∈Ω), update the λ value of non-boundary positions; Step 306: For a given convergence criterion ε>0, check the criterion ‖λ point by point. n+1 -λ n ‖ L1 ≤ε to check convergence; Step 307: When the convergence is satisfied, the corresponding The value is used as the gradient of the objective function; in, is the normal vector perpendicular to the surface of the vibration pickup sensor, is the gradient, T is the calculated arrival time of the transmitting signal source, λ is the Lagrange multiplier, T° refers to the propagation time of the longitudinal and / or shear waves of the vibration waves observed by the vibration pickup sensor, Ω is the vibration pickup sensor surface, x∈Ω refers to the position x on the vibration pickup sensor surface, ‖λ n+1 -λ n ‖ L1 is the Lagrange multiplier matrix L 1 norm, indicating that the algorithm converges when the rate of change of the Lagrange multiplier is negative, λ n+1 is the λ value obtained in the n+1th iteration, λ n is the value of λ obtained at the nth iteration.

4. The method of inversion based on active and passive dual-source combined shock wave CT according to claim 3 is characterized in that: In step 4, the steps to output the final source parameters and wave velocity are: When the shock wave travels, the residual E and RMS residual If a stable state is reached, the inversion is judged to be converged, and the final source parameters and wave velocity are output, where N is the number of observed vibrations; otherwise, step 302 is entered for iteration.

5. The method for inversion of active and passive dual-source shock wave CT according to claim 2 is characterized in that: The probabilistic non-uniform initial velocity ranges from 1000m / s to 6500m / s.

6. A rock burst warning method, characterized in that: The inversion model obtained by the active-passive dual-source combined shock wave CT inversion method as described in any one of claims 1 to 5 is used for rock burst warning.

Images