Pre-splitting blasting damage analysis method and system based on eccentric decoupling charge structure

Abstract

The application discloses a pre-splitting blasting damage analysis method and system based on an eccentric decoupling charge structure, relates to the technical field of blasting engineering, and comprises the following steps: obtaining rock mass mechanics and blasting parameters, constructing a numerical model and setting a safety threshold; establishing an eccentric decoupling charge model, calculating a rock mass theoretical damage field in a forward direction by precisely applying an asymmetric initial load; using a virtual detection wave emitted by inter-hole micro-difference delay to inversely calculate a real damage field and dynamically correct the rock mass elastic modulus; fusing the theoretical and real damage field depths, reconstructing an asymmetric damage field and outputting a three-dimensional tomography cloud chart containing confidence; and checking the cloud chart features and the safety threshold, iteratively recalculating in a closed loop and reversely outputting optimal blasting parameter configurations. The application realizes accurate fusion of theoretical calculation and field measurement data under complex eccentric charge working conditions, and constructs a blasting parameter closed-loop automatic optimization system which takes into account safety vibration control and efficient fragmentation.

Description

Technical Field

[0001] This invention relates to the field of blasting engineering technology, and in particular to a method and system for analyzing pre-splitting blasting damage based on an eccentrically decoupled charge structure. Background Technology

[0002] In geotechnical engineering and mining, pre-splitting blasting is often used to protect and preserve the integrity of rock masses. In actual construction, due to gravity or construction conditions, the explosive charge inside the borehole often adheres to one side of the borehole wall, forming an eccentric, uncoupled charge structure. However, existing numerical simulation methods for blasting typically simplify the explosive load as a symmetrical pressure uniformly distributed along the borehole wall, severely deviating from the actual asymmetric stress state and leading to distortion of the initial calculation boundaries.

[0003] Furthermore, rock masses experience cumulative damage under continuous blasting, leading to a dynamic deterioration of their mechanical parameters. However, traditional simulations often employ static parameter evolution models, lacking a mechanism to capture real physical damage during blast intervals and make real-time corrections, resulting in the continuous accumulation of errors. Moreover, the blasting parameter design phase heavily relies on manual experience and trial-and-error based on diagrams, lacking multi-objective, fully automated closed-loop optimization capabilities that balance safe vibration velocity control with fracture effects, making it difficult to meet the demands of refined and intelligent blasting operations under complex conditions. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a pre-splitting blasting damage analysis method based on an eccentric uncoupled charge structure, which solves the problems of distorted blasting damage simulation boundaries, lack of real-time error correction in dynamic evolution, and high dependence on manual experience and guesswork in parameter design under complex eccentric charge conditions.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a pre-splitting blasting damage analysis method based on an eccentric uncoupled charge structure, which includes obtaining rock mass physical and mechanical parameters and initial blasting design parameters, constructing a rock mass numerical model, and setting a safe vibration velocity threshold. Based on the eccentric uncoupled charge geometric model constructed from the initial blasting design parameters, the asymmetric initial borehole wall pressure is calculated as the initial load, non-uniform borehole wall boundary loading is performed, and forward calculation is performed based on the initial load to obtain the theoretical damage field of the rock mass. Virtual probe waves are emitted within a small time difference to perform inversion, obtain the real damage field of the rock mass, and dynamically correct the elastic modulus of the rock mass in real time. The theoretical damage field is fused with the real damage field to reconstruct the asymmetric damage field and output a three-dimensional tomographic cloud map containing confidence level. Extract the boundary peak vibration velocity and fracture state from the three-dimensional tomographic cloud map, verify them with the safe vibration velocity threshold, adjust the blasting parameters in reverse and recalculate in closed loop, and output the optimal parameter configuration. The three-dimensional tomographic cloud map is a visualized three-dimensional data field generated by jointly visually encoding the reconstructed damage variables and data inversion reliability based on the spatial coordinates of the rock mass.

[0007] As a preferred embodiment of the pre-splitting blasting damage analysis method based on the eccentric uncoupled charge structure described in this invention, the rock mass physical and mechanical parameters include density, compressive strength, tensile strength, elastic modulus, and initial longitudinal wave velocity; the initial blasting design parameters include borehole diameter, charge diameter, single-hole charge quantity, decoupling coefficient, and inter-hole differential delay time.

[0008] As a preferred embodiment of the pre-splitting blasting damage analysis method based on the eccentric uncoupled charge structure described in this invention, the non-uniform hole wall boundary loading includes establishing a polar coordinate system with the hole center as the pole and defining the central angle. Based on the aforementioned eccentric uncoupled charge geometric model, the radial expansion distance of the explosive gas from the surface of the charge cartridge to the inner wall of the borehole is calculated at different central angles. Based on the radial expansion distance, the inner wall of the borehole is divided into a wall-attached contact area and a cavity expansion area; The wall-attached contact area is the area where the outer surface of the explosive charge is in direct contact with the inner wall of the borehole, and the explosive shock wave is directly coupled without being attenuated by the gap medium. The initial peak pressure generated by the explosive detonation is directly used as the borehole pressure of the wall-attached contact area. The cavity expansion zone is the area of ​​the borehole wall where there is an initial spatial gap between the outer surface of the explosive cartridge and the inner wall of the borehole, and which is subjected to the effect of the explosive gas after adiabatic expansion and attenuation through the gap. The radial expansion distance is substituted into the adiabatic expansion equation of the explosive gas to calculate the borehole wall pressure after attenuation by the medium. By combining the calculation results of the contact zone with the cavity expansion zone, an asymmetric initial borehole wall pressure distribution function that continuously varies with the central angle is constructed. The asymmetric initial borehole wall pressure distribution function is then mapped and transformed into the mechanical boundary conditions of the borehole wall grid nodes in the rock mass numerical model, thus completing the non-uniform loading.

[0009] As a preferred embodiment of the pre-splitting blasting damage analysis method based on the eccentric uncoupled charge structure described in this invention, the real damage field is a physical damage distribution reconstructed by inversion based on the virtual probe wave reflection signal, which is used to drive the dynamic correction of the rock mass elastic modulus in real time. Extract the differential delay time between adjacent boreholes from the initial blasting design parameters, set the differential delay time as the virtual detection time window in the numerical simulation, and pause the application of new explosive loads; Within the virtual detection time window, a virtual broadband detection wave is generated at the preset observation node in the rock mass numerical model to penetrate the rock mass area that has been affected by the previous blasting. Extract the transmitted and reflected wave signals of the probe wave after it passes through the damaged rock mass, and quantitatively calculate the wave velocity reduction rate and amplitude attenuation coefficient. By combining the full waveform inversion algorithm, the wave velocity reduction rate and amplitude attenuation coefficient are mapped and transformed into real damage variables with spatial grid distribution, thereby obtaining the real damage field of the rock mass; Based on the actual damage field, the elastic modulus of each grid cell in the rock mass numerical model is covered and dynamically corrected one by one using the elastic modulus degradation equation, and the corrected rock mass numerical model is used as the loading medium for the next micro-delay detonation.

[0010] As a preferred embodiment of the pre-splitting blasting damage analysis method based on the eccentric uncoupled charge structure described in this invention, wherein: the reconstructing of the asymmetric damage field includes constructing a generative adversarial network model containing a generator network and a discriminator network; The theoretical damage field, the actual damage field, and the spatial coordinates of the grid cells in the rock mass numerical model are extracted and spliced ​​into tensors. The spliced ​​tensors are then input into the generator network to output the predicted damage field. The theoretical damage field and the actual damage field are used as theoretical samples and measured samples, respectively, and input with the predicted damage field into the discriminator network to calculate and output the discrimination probability of the predicted damage field; A composite loss function is constructed, which includes adversarial loss, data fidelity loss, and physical mechanism loss. The adversarial loss is the deviation between the discrimination probability and a preset threshold. The data fidelity loss is the calculation error between the predicted damage field and the real damage field at the virtual detection node. The physical mechanism loss is the calculation residual obtained by substituting the spatial gradient of the predicted damage field into the rock mass wave equation. The generator network is trained alternately using the composite loss function until convergence, and the output of the generator network after convergence is used as the asymmetric damage field. The measured data obtained after the explosion are updated to update the measured samples, and the generative adversarial network model is incrementally trained.

[0011] As a preferred embodiment of the pre-splitting blasting damage analysis method based on the eccentric uncoupled charge structure described in this invention, the output of the three-dimensional tomographic cloud map containing confidence includes obtaining the spatial coordinates and damage variables of each grid cell in the asymmetric damage field, and simultaneously extracting the data confidence parameters corresponding to each grid cell. Construct a multidimensional visual mapping rule to map the damage variable to a first visual attribute and the data confidence parameter to a second visual attribute; Perform three-dimensional spatial interpolation based on the spatial coordinates of each grid cell, the first visual attribute, and the second visual attribute; The interpolated spatial data is reconstructed using a 3D volume rendering algorithm to output the 3D tomographic cloud map containing confidence level.

[0012] As a preferred embodiment of the pre-splitting blasting damage analysis method based on the eccentric uncoupled charge structure described in this invention, the reverse adjustment of blasting parameters and closed-loop recalculation includes obtaining the peak vibration velocity distribution data of the three-dimensional tomographic cloud map at the preset protection boundary, as well as the effective damage volume within the target blasting area. A preset target vibration velocity distribution is introduced, and the distribution distance measure between the peak vibration velocity distribution data and the target vibration velocity distribution is calculated. A multi-objective fitness function F is constructed with the goal of minimizing the distribution distance measure and maximizing the effective damage volume. The hard constraint condition for the multi-objective fitness function F is set as the peak velocity distribution data. The maximum value in the range is less than or equal to the safe vibration velocity threshold. ; Based on the fitness function F and the hard constraints, the initial blasting design parameter set is updated using an optimization algorithm to obtain the new blasting parameter set for the current iteration. The new blasting parameter set is imported into the rock mass numerical model and the steps of forward blasting loading, virtual probe wave inversion and asymmetric damage field reconstruction are re-executed to generate an updated three-dimensional tomographic cloud map. The variable update and reconstruction steps are executed repeatedly until the fitness function converges and satisfies the hard constraint condition. The new burst parameter set in the current convergent state is then extracted as the optimal parameter configuration.

[0013] Secondly, the present invention provides a pre-splitting blasting damage analysis system based on an eccentric uncoupled charge structure, including a data acquisition unit for acquiring rock mass physical and mechanical parameters and initial blasting design parameters, constructing a rock mass numerical model, and setting a safe vibration velocity threshold. The deduction unit, based on the eccentric uncoupled charge geometric model constructed from the initial blasting design parameters, calculates the asymmetric initial borehole wall pressure as the initial load, performs non-uniform borehole wall boundary loading, and performs forward calculation based on the initial load to obtain the theoretical damage field of the rock mass. The inversion unit emits virtual probe waves within a micro-difference delay time to perform inversion, obtain the real damage field of the rock mass, and dynamically correct the elastic modulus of the rock mass in real time. The fusion unit fuses the theoretical damage field with the real damage field to reconstruct the asymmetric damage field and outputs a three-dimensional tomographic cloud map containing confidence level. The optimization unit extracts the boundary peak vibration velocity and crack state from the three-dimensional tomographic cloud map, verifies them with the safe vibration velocity threshold, adjusts the blasting parameters in reverse and recalculates in a closed loop, and outputs the optimal parameter configuration.

[0014] Thirdly, the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, it implements any step of the pre-splitting blasting damage analysis method based on an eccentrically decoupled charge structure as described in the first aspect of the present invention.

[0015] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the pre-splitting blasting damage analysis method based on an eccentrically decoupled charge structure as described in the first aspect of the present invention.

[0016] The beneficial effects of this invention are as follows: A numerical model is constructed by acquiring rock mass and blasting parameters and setting a safety threshold; an eccentric charge model is established, asymmetric loads are applied, and the theoretical damage field is calculated in the forward direction; the real damage field is inverted using virtual probe waves emitted with a small time delay, and the elastic modulus is dynamically corrected; the theoretical and real damage fields are fused, and a three-dimensional tomographic cloud map with confidence level is reconstructed and output; cloud map features are extracted and verified with the safety threshold, and after closed-loop iterative recalculation, the optimal blasting parameter configuration is finally output in reverse. This achieves high-fidelity deduction and real-time dynamic error correction of blasting damage evolution under complex eccentric charge conditions, and establishes a fully automatic closed-loop optimization link for blasting parameters based on three-dimensional tomographic cloud maps, effectively overcoming the problems of traditional simulation boundary distortion and the high dependence on manual guesswork in scheme design. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. 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 these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a pre-splitting blasting damage analysis method based on an eccentrically decoupled charge structure.

[0019] Figure 2 This is a diagram of a computer device used for a pre-splitting blasting damage analysis method based on an eccentrically decoupled charge structure. Detailed Implementation

[0020] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0021] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0022] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0023] Reference Figures 1-2 This is one embodiment of the present invention, which provides a pre-splitting blasting damage analysis method based on an eccentrically decoupled charge structure, including the following steps: S1: Obtain the physical and mechanical parameters of the rock mass and the initial blasting design parameters, construct the numerical model of the rock mass, and set the safe vibration velocity threshold.

[0024] To achieve high-fidelity digital reconstruction of the real geological environment and to provide the underlying physical space and rigid safety constraints for subsequent asymmetric loading simulation and parameter closed-loop optimization, a three-dimensional digital target that closely matches the real geological conditions is reconstructed in the virtual computing space by accurately collecting on-site rock mechanics and blasting design data.

[0025] The rock mass physical parameters of the target blasting area are obtained through on-site geological exploration (such as sonic logging technology) and indoor rock mechanics tests. These rock mass physical and mechanical parameters include density, compressive strength, tensile strength, elastic modulus, and initial longitudinal wave velocity, which serve as the core basic data for defining the constitutive relationship of the rock mass material in subsequent numerical simulations.

[0026] Based on the preliminary engineering construction plan at the site, initial blasting design parameters were extracted. These initial blasting design parameters include borehole diameter, explosive cartridge diameter, single-hole explosive charge, decoupling coefficient, and inter-hole differential delay time, which determine the total input scale of the explosion energy and the energy release pattern over time.

[0027] Based on the borehole diameter, borehole spacing, and row spacing parameters in the initial blasting design parameters, a three-dimensional geometric model including the borehole cavity and the surrounding retained rock mass is constructed. Specifically, an initial three-dimensional rock mass matrix entity is first generated according to the macroscopic boundary dimensions of the calculation region. Using the detonation origin as a reference, the central axis of each borehole is arrayed and positioned in the three-dimensional coordinate system using the borehole spacing and row spacing parameters. Then, the corresponding cylindrical borehole entity is generated by stretching along the borehole depth direction using the borehole diameter as a feature dimension. Using a three-dimensional geometry engine, Boolean difference operations are performed to accurately excavate the cylindrical borehole entities from the initial three-dimensional rock mass matrix entity, thereby defining the hollow borehole cavity. Finally, a three-dimensional geometric model including the borehole cavity and the surrounding retained rock mass is output.

[0028] The three-dimensional geometric model is spatially discretized and meshed. Specifically, a locally refined mesh is used in the loaded region near the borehole wall, while a gradually sparser mesh is used in regions farther from the borehole, transforming the continuous geometry into a finite number of mesh elements. A dynamic damage constitutive relation for the rock is introduced, and the acquired physical and mechanical parameters of the rock mass are input into the mesh elements, assigning dynamic computational properties to the initial rock mass mesh model.

[0029] By applying non-reflective transmission boundary conditions to the outer truncated surface of the overall grid, the non-realistic reflection of stress waves on the artificial truncated boundary is eliminated, thereby completing the construction of the numerical model of the rock mass.

[0030] The system acquires attribute information of objects to be protected around the target blasting area. Based on the rock mass numerical model, it accurately calculates the shortest spatial straight-line distance from the foundation boundary of each object to be protected to the preset blasting vibration source center, i.e., the spatial safety distance. The attribute information of the objects to be protected includes: building (structure) type, importance level, structural service status, and seismic fortification level (e.g., general civil buildings, existing operational tunnels, or steep slopes).

[0031] Combining the attribute information of the object to be protected with the estimated dominant frequency band of blasting vibration (such as low frequency, medium frequency, or high frequency), and in accordance with the vibration safety allowable standards specified in the "Blasting Safety Regulations" (GB6722), the permissible upper limit of peak particle vibration velocity acting on the foundation of the object to be protected is determined through multi-dimensional table lookup and interpolation algorithms. This permissible upper limit of peak particle vibration velocity is set as the safe vibration velocity threshold, serving as a hard safety constraint boundary in the subsequent reverse adjustment of blasting parameters and closed-loop recalculation.

[0032] S2: Based on the eccentric uncoupled charge geometric model constructed from the initial blasting design parameters, the asymmetric initial borehole wall pressure is calculated as the initial load, non-uniform borehole wall boundary loading is performed, and forward calculation is performed based on the initial load to obtain the theoretical damage field of the rock mass.

[0033] To break the idealized uniform and symmetrical loading assumption in traditional numerical simulations, and to realistically reproduce the asymmetrical stress state of the drug cartridges eccentrically attached to the wall due to gravity or construction tilt at the engineering site, this study provides a scientific and accurate mechanical prior benchmark for subsequent dynamic measurement inversion and multi-source data fusion.

[0034] Extract the explosive charge diameter and preset eccentricity (usually expressed as polar angle in a two-dimensional cross-sectional polar coordinate system) from the initial blasting design parameters. In the constructed borehole cavity, generate an initial explosive charge entity with the explosive charge diameter as the characteristic dimension, so that the central axis of the initial explosive charge entity coincides with the central axis of the borehole cavity.

[0035] In practical engineering, the propellant cartridge will inevitably become eccentric due to gravity and the roughness of the borehole wall. Based on the difference between the borehole diameter and the propellant cartridge diameter, the maximum unidirectional radial movement allowance of the propellant cartridge within the borehole is calculated, i.e., (borehole diameter - propellant cartridge diameter) / 2. Combining this with the preset eccentric orientation as the vector direction, the eccentric translation vector of the initial propellant cartridge entity is calculated in the coordinate system.

[0036] The initial charge body is spatially translated according to the eccentric translation vector, so that the outer surface and the inner wall of the borehole cavity are in contact at the preset eccentric orientation, thereby defining the asymmetric annular gap and completing the establishment of the eccentric uncoupled charge geometric model.

[0037] Establish a polar coordinate system with the center of the borehole as the pole, define the central angle, and use the preset eccentric orientation as the polar axis (i.e., polar angle). =0°), establish a polar coordinate system on the cross-section of the borehole. Based on the eccentric uncoupled charge geometry model, calculate the radial expansion distance of the explosive gas from the surface of the charge cartridge to the inner wall of the borehole at different central angles. According to the radial expansion distance, the inner wall of the borehole is divided into a wall-attached contact area and a cavity expansion area.

[0038] The contact area between the outer surface of the explosive charge and the inner wall of the borehole is the region of the borehole wall where the explosive shock wave is directly coupled without being attenuated by the interstitial medium, thus amplifying the initial peak pressure generated by the explosive detonation. The pressure is directly used as the borehole wall pressure in the contact area with the wall. The cavity expansion area is the borehole wall region where there is an initial spatial gap between the outer surface of the explosive cartridge and the inner wall of the borehole, and which is subjected to the effect of the explosive gas after adiabatic expansion and attenuation through the gap. The radial expansion distance is substituted into the adiabatic expansion equation of the explosive gas to calculate the borehole wall pressure after attenuation by the medium.

[0039] By combining the calculation results of the wall-attached contact area and the cavity expansion area, an asymmetric initial borehole wall pressure distribution function that continuously varies with the central angle is constructed. , can be represented as: in, This represents the asymmetric initial borehole wall pressure distribution function; This indicates the initial peak pressure generated by the detonation of the explosive; This represents the critical central angle that divides the wall-attached contact area from the cavity expansion area; Indicates the eccentricity of the medicine roll; Indicates the radius of the medicine roll; Indicates the borehole radius; This indicates the adiabatic index of the explosive gas.

[0040] The asymmetric initial borehole wall pressure distribution function is mapped and transformed into the mechanical boundary conditions of the borehole wall grid nodes in the rock mass numerical model, thus completing the non-uniform loading.

[0041] Activate the explicit dynamics solver in the numerical computation platform. Then, construct the asymmetric initial borehole wall pressure distribution function. Combining the explosion load time history decay curve (such as an exponential decay function), a transient mechanical boundary condition with dual coupling of time and space (polar angle) is transformed and applied to the grid nodes of the borehole wall. Simultaneously, based on the characteristic size of the grid elements and the wave velocity of the rock mass, an explicit time integration step size in the microsecond range is automatically calculated and set according to the Courant-Friedrich-Lyuvi conditions (CFL conditions). We initiate a central differential time integral solution to ensure the numerical stability of the solution for the physical process of the explosive impact.

[0042] Within each time integration step, the governing equations of continuum dynamics based on Newton's second law are solved. The transmission, reflection, and dynamic propagation of explosion stress waves (including longitudinal, transverse, and surface waves) in each grid cell of the rock mass numerical model are accurately tracked. By monitoring the nodal displacements and velocities of each grid cell, the transient stress tensors of each grid cell are derived and obtained in real time. (Including compressive stress, tensile stress, and shear stress components) and transient strain rate .

[0043] The obtained transient stress tensor With transient strain rate The data is input in real time into the pre-defined dynamic damage constitutive relation of the rock. Based on the input strain rate... The yield surface and tensile strength of the rock are dynamically adjusted (i.e., strain rate hardening effect); when the transient stress tensor exceeds the dynamic yield limit or tensile strength, the damage evolution equation is triggered, and the cumulative physical damage variables of each mesh element are calculated and updated. .in, It is a dimensionless scalar, and The rock mass is intact. Characterizes complete fracture and failure of the rock mass.

[0044] Once the set total solution time is reached (i.e., the blasting stress wave energy decays to near zero), the forward dynamic calculation ends. The data post-processing interface is then invoked to traverse the entire mesh in the rock mass numerical model and extract the... The spatial centroid coordinates of each grid cell and the final cumulative physical damage variable corresponding to the end of the blast. All extracted grid cell data are structurally stitched together to construct a three-dimensional spatial dataset containing global location and damage status. The dataset can then serve as the basis for visualization and subsequent fusion analysis, generating the theoretical damage field.

[0045] The theoretical damage field refers to the idealized physical state of the rock mass under the initial action of explosive load, which is obtained through numerical calculation based on rock dynamics theory and blasting mechanics model, reflecting the distribution of internal damage variables with spatial coordinates.

[0046] It achieves a high-fidelity physical reconstruction of the energy release process of asymmetric blasting under real eccentric conditions, and derives a pure theoretical rock mass damage field benchmark that has not been measured and corrected.

[0047] S3: During the micro-delay time, a virtual probe wave is emitted for inversion to obtain the real damage field of the rock mass and to dynamically correct the elastic modulus of the rock mass in real time.

[0048] To achieve dynamic experimental error correction of theoretical simulation errors and real-time adaptive updating of the mechanical properties of the loaded medium, this method breaks through the limitations of traditional one-way numerical simulation.

[0049] The differential delay time between adjacent boreholes in the initial blasting design parameters is extracted and set as the virtual detection time window in the numerical simulation, pausing the application of new blasting loads. Within the virtual detection time window, a virtual broadband probe wave is excited at a preset observation node in the rock mass numerical model, penetrating the rock mass area already affected by previous blasting. The transmitted and reflected wave signals after the probe wave passes through the damaged rock mass are extracted, and the wave velocity reduction rate and amplitude attenuation coefficient are quantitatively calculated.

[0050] Specifically, in the rock mass numerical model, a spatial array of virtual source nodes and detector nodes is deployed around the rock mass area damaged by previous blasting. A broadband transient dynamic pulse function (such as a Ricker wavelet) is selected as the probe source, which is converted into a transient load or displacement boundary condition and applied to the virtual source node, activating the explicit dynamic solver. Within a preset differential detection time window, the generated virtual broadband probe wave is driven to propagate forward within the model, forcibly penetrating the grid cell region with accumulated physical damage variables (D>0). Since the probe wave will experience a sudden change in wave impedance and energy dissipation when encountering damaged grids, high-frequency sampling is performed at the detector nodes to derive the full-echo transmission and reflection waveform signal containing physical scattering and attenuation characteristics, accurately extracting the first arrival time of the transmitted wave. and peak waveform amplitude Simultaneously, the initial excitation wave initiation time at the virtual seismic source node was retrieved. With the initial peak amplitude .

[0051] Calculate the actual equivalent propagation velocity of the probe wave in the damaged rock mass. , can be represented as: in, This represents the actual equivalent wave velocity of the probe wave after it passes through the damaged rock mass; This represents the straight-line propagation distance between the virtual source node and the detector node; Indicates the initial moment of the virtual seismic source's activation; This indicates the first arrival time of the transmitted wave captured by the detector node.

[0052] Retrieve the initial longitudinal wave velocity from the physical and mechanical parameters of the rock mass. (i.e., the reference wave velocity under undamaged rock mass conditions), the wave velocity reduction rate is calculated. , can be represented as: in, Indicates the rate of decrease in wave velocity; This represents the reference wave velocity of the rock mass in its initial undamaged state; This represents the actual equivalent wave velocity.

[0053] Based on the exponential energy decay law of explosive stress waves in viscoelastic damaged media, the initial peak amplitude of the virtual source is utilized. The peak amplitude of transmission / reflection received by the detector node Calculate the amplitude attenuation coefficient , can be represented as: in, Indicates the amplitude attenuation coefficient; This represents the straight-line propagation distance between the virtual source node and the detector node; This represents the initial peak amplitude excited by the virtual source; This represents the peak amplitude of the transmitted wave received by the detector node.

[0054] By constructing a reduction rate based on measured wave velocity With amplitude attenuation coefficient Given the hard-constrained objective residual function, a simultaneous full-waveform inversion algorithm is used to calculate the sensitivity kernel function (i.e., gradient direction) of the objective function relative to the damage variables of all grid cells in the global domain. Under a preset convergence criterion, a numerical optimization algorithm (such as L-BFGS or the conjugate gradient method) is applied to the damage variables of each spatial grid cell. By performing successive corrections and dynamic optimization, the wave velocity reduction rate and amplitude attenuation coefficient are mapped and transformed into real damage variables with a spatial grid distribution, thereby obtaining the real damage field of the rock mass. It contains a four-dimensional tensor set (i.e., a set of real rock mass damage field data) containing spatial coordinates and inverted damage values. , For the first The actual damage variable values ​​of each grid cell after inversion correction.

[0055] It should be noted that the full waveform inversion algorithm, as a high-resolution parameter reconstruction technique, first uses a spatial rock mass mesh with theoretical damage variables as the initial background model, and then uses a dynamic finite element solver to perform a forward wavefield simulation to obtain the theoretically calculated waveform records at each receiver node. The theoretically calculated waveform records are then compared with those obtained from on-site virtual detection, which include the wave velocity reduction rate. With amplitude attenuation coefficient The measured waveform signals of attenuation characteristics are used for wavefield residual matching to construct a residual objective function based on waveform difference. The adjoint state method is introduced to calculate the gradient (i.e., sensitivity kernel function) of the residual objective function with respect to the elastic parameters of each grid cell in the global domain, to determine the update direction of the model parameters. Finally, nonlinear optimization algorithms such as L-BFGS are used to iteratively correct and dynamically optimize the mechanical parameters of each spatial grid along the gradient descent direction until the residual objective function converges to a preset threshold. Through inversion closure, the measured wave attenuation characteristics are accurately mapped and transformed into the real damage field dataset. .

[0056] The real damage field is the physical damage distribution reconstructed from the reflected signal of the virtual probe wave.

[0057] Based on the actual damage field, the elastic modulus of each grid cell in the rock mass numerical model is covered and dynamically corrected one by one using the elastic modulus degradation equation, and the corrected rock mass numerical model is used as the loading medium for the next micro-delay detonation.

[0058] Specifically, by automatically parsing and traversing the acquired set of real rock mass damage field data, the real physical damage variables corresponding to each spatial grid cell in the rock mass numerical model are extracted. Using the elastic modulus degradation equation based on the theory of continuous damage mechanics ( ,in This represents the initial undamaged elastic modulus of the rock mass. (Representing the effective elastic modulus after degradation), the effective stiffness parameters of each mesh element under the current damaged state are calculated in real time. The calculated effective elastic modulus of each element is then used... In the material property matrix of the numerical solver, the original constitutive parameters are precisely replaced one by one at the spatial coordinate level and fully covered to complete the dynamic correction of macroscopic mechanical properties. Finally, the numerical model of the rock mass that has undergone parameter reconstruction and mechanical degradation is directly used as the initial physical environment and loading medium for the next micro-delay blasting borehole initiation, thus realistically reproducing the propagation and nonlinear attenuation law of the subsequent explosion stress wave in the deteriorated rock mass, and completely closing the loop of the dynamic digital twin simulation process of multi-hole micro-delay blasting.

[0059] It achieves dynamic closed-loop error correction of simulation errors and adaptive updating of rock mass mechanical properties, enabling numerical simulation to completely leap from traditional unidirectional static calculation to digital twins that closely approximate real working conditions.

[0060] S4: Fuse the theoretical damage field with the real damage field to reconstruct the asymmetric damage field and output a three-dimensional tomographic cloud map containing confidence level.

[0061] To achieve deep intelligent fusion of prior physical mechanisms and measured inversion data, and to eliminate the cumulative errors of single theoretical deductions and the spatial blind spots of physical detection.

[0062] A generative adversarial network (GAN) model comprising a generator network and a discriminator network is constructed. Specifically, due to the three-dimensional spatial characteristics of the rock mass damage field, the generator network adopts an encoder-decoder architecture based on a 3D U-Net, extracting high-dimensional features of spatial centroid coordinates and damage variables through 3D convolutional layers and preserving spatial details through skip connections; the discriminator network adopts a fully convolutional network (FCN) architecture. The spatial coordinates of the theoretical damage field, the actual damage field, and the grid cells in the rock mass numerical model are extracted, and tensors are concatenated. The concatenated tensor is then input into the generator network to output the predicted damage field. .

[0063] The theoretical damage field and the actual damage field are used as theoretical samples and measured samples, respectively, and input into the discriminator network along with the predicted damage field to calculate and output the discrimination probability of the predicted damage field.

[0064] Furthermore, by using the theoretical damage field as a spatial condition constraint tensor, and simultaneously inputting it, along with the real damage field and the predicted damage field output by the generator, into a discriminator network based on a three-dimensional convolutional architecture, the network is used to extract high-dimensional feature differences in the spatial fracture distribution of multi-source data, thereby calculating and outputting a discriminant probability value for quantitatively evaluating how closely the predicted field approximates the real field.

[0065] Build with adversarial loss Data fidelity loss and physical mechanism loss Composite loss function , can be represented as: in, Represents the composite loss function; Indicates resistance to loss; Indicates the weighting coefficient against losses; This indicates a loss in data fidelity. Weighting coefficients representing data fidelity loss; Indicates loss due to physical mechanisms; The weighting coefficients represent the physical mechanism loss.

[0066] The adversarial loss is the deviation between the discrimination probability and the preset threshold; the data fidelity loss is the calculation error between the predicted damage field at the virtual detection node and the real damage field; the physical mechanism loss is the calculation residual obtained by substituting the spatial gradient of the predicted damage field into the rock mass wave equation.

[0067] The generator network is trained alternately using the composite loss function until convergence, and the output of the generator network after convergence is used as the asymmetric damage field.

[0068] The asymmetric damage field is a three-dimensional physical state in which the spatial degree of fracture and mechanical deterioration characteristics are asymmetrically distributed due to the uneven release of stress wave energy within the rock mass medium under non-uniform (such as eccentric wall-mounted) blasting loads.

[0069] The measured data obtained after the explosion are updated to update the measured samples, and the generative adversarial network model is incrementally trained.

[0070] After the generative adversarial network model completes incremental iteration and dynamic convergence, it outputs the final fused and reconstructed asymmetric damage field underlying data set. To transform the high-dimensional tensor data into visual evidence for intuitive decision-making in engineering settings, further data deconstruction and rendering are performed.

[0071] The data confidence parameters corresponding to each grid cell are obtained by using Monte Carlo random deactivation technology. During the inference phase of the generative adversarial network model, the measured samples are propagated multiple times to calculate the statistical variance of the predicted damage variable of each grid cell in the multiple inference results. The statistical variance is then normalized and inverted to be transformed into a confidence parameter for evaluating the reliability of data inversion. The smaller the variance, the higher the confidence.

[0072] Obtain the spatial coordinates and damage variables of each grid element in the asymmetric damage field. Simultaneously extract the data confidence parameters (i.e., the reliability probability of the predicted data for each grid cell) corresponding to each grid cell.

[0073] Construct a multidimensional visual mapping rule to map the impairment variable to a first visual attribute (e.g., color hue, set). The undamaged area is a cool color such as blue, which gradually changes to a high warning warm color such as red as the damage variable increases, representing the severely shattered area. The data confidence parameter is mapped to a second visual attribute (such as the transparency channel Alpha or brightness, setting the high confidence area to be an opaque entity, and the low confidence area to be presented in a semi-transparent or blurred state).

[0074] Three-dimensional spatial interpolation is performed based on the spatial coordinates of each grid cell, the first visual attribute, and the second visual attribute. Here, the spatial coordinates are treated as an independent one-dimensional basic attribute dimension for global distance weighting. Due to the physical constraints of the explosion stress wave attenuation mechanism, the internal damage of the rock mass exhibits a continuous and smooth gradual evolution in real space. Interpolation can eliminate the step effect caused by the discrete grid and accurately extrapolate the attributes of unsampled areas.

[0075] A 3D volume rendering algorithm is used to reconstruct the internal perspective of the interpolated spatial data, outputting a 3D tomographic cloud map containing confidence levels. Specifically, a ray casting algorithm is used to sample the interpolated voxel field at equal intervals along a virtual line of sight. The damage variable of the sampled points is mapped to RGB color using the optical transfer function, and the confidence parameter is mapped to alpha opacity. Finally, the optical properties of each point are accumulated and synthesized along the ray, and a 3D tomographic cloud map with both damage color and confidence transparency features is rendered.

[0076] The three-dimensional tomographic cloud map is a visualized three-dimensional data field generated by jointly visually encoding the reconstructed damage variables and data inversion reliability based on the spatial coordinates of the rock mass.

[0077] It achieves high-dimensional tensor reconstruction that combines the physical rigor of numerical simulation with the spatial realism of field measurements. By assigning clear probabilistic evaluation labels to invisible fracture zones, it provides a transparent and quantifiable three-dimensional decision-making foundation for the dynamic optimization of blasting schemes.

[0078] S5: Extract the boundary peak vibration velocity and crack state from the three-dimensional tomographic cloud map, verify them with the safe vibration velocity threshold, adjust the blasting parameters in reverse and recalculate in closed loop, and output the optimal parameter configuration.

[0079] Obtain the peak velocity distribution data V of the three-dimensional tomographic cloud image at the preset protective boundary. ppv Given the effective damage volume within the target blasting area, a preset target vibration velocity distribution is introduced. The distribution distance measure between the peak vibration velocity distribution data and the target vibration velocity distribution is calculated (the spatial mean square error or Wasserstein distance can be used as the distribution distance measure). A multi-objective fitness function F is constructed with the goal of minimizing the distribution distance measure and maximizing the effective damage volume.

[0080] The hard constraint condition for the multi-objective fitness function F is set as the peak velocity distribution data. The maximum value in the range is less than or equal to the safe vibration velocity threshold. Based on the fitness function and the hard constraints, an optimization algorithm (such as the non-dominated sorting genetic algorithm NSGA-II or the multi-objective particle swarm optimization algorithm MOPSO) is used to update the variables of the initial blasting design parameter set to obtain the new blasting parameter set for the current iteration.

[0081] The new blasting parameter set is imported into the rock mass numerical model and the steps of forward blasting loading, virtual probe wave inversion, and asymmetric damage field reconstruction are re-executed to generate an updated three-dimensional tomographic cloud map.

[0082] Specifically, the newly acquired blasting parameter set is reloaded into the rock mass numerical model as the initial boundary condition. The dynamic finite element calculation engine is triggered to simulate the stress wave propagation and initial damage evolution under the blast load. Then, within a preset time step, the virtual acoustic wave detection algorithm is automatically invoked to perform secondary perspective analysis on the damaged medium. The obtained detection vector and physical and mechanical residuals are synchronously fed back to the pre-trained generative adversarial network for asymmetric damage field feature reconstruction. Thus, a high-fidelity three-dimensional tomographic cloud map reflecting the true effectiveness of the current design scheme is dynamically mapped and generated in a unified spatial coordinate system.

[0083] The variable update and reconstruction steps are executed repeatedly until the fitness function converges and satisfies the hard constraint condition. The new burst parameter set in the current convergent state is then extracted as the optimal parameter configuration.

[0084] It achieves automatic adjustment of blasting design parameters to balance rock breaking volume and vibration intensity by using a safe vibration velocity threshold as the performance boundary, thereby constructing a closed-loop control link from damage field assessment to automatic optimization of blasting scheme.

[0085] This embodiment also provides a pre-splitting blasting damage analysis system based on an eccentric uncoupled charge structure, including: a data acquisition unit, which acquires rock mass physical and mechanical parameters and initial blasting design parameters, constructs a rock mass numerical model, and sets a safe vibration velocity threshold.

[0086] The deduction unit, based on the eccentric uncoupled charge geometric model constructed from the initial blasting design parameters, calculates the asymmetric initial borehole wall pressure as the initial load, performs non-uniform borehole wall boundary loading, and performs forward calculations based on the initial load to obtain the theoretical damage field of the rock mass.

[0087] The inversion unit emits virtual probe waves within a micro-difference delay time to perform inversion, obtain the true damage field of the rock mass, and dynamically correct the elastic modulus of the rock mass in real time.

[0088] The fusion unit fuses the theoretical damage field with the real damage field to reconstruct the asymmetric damage field and outputs a three-dimensional tomographic cloud map containing confidence levels.

[0089] The optimization unit extracts the boundary peak vibration velocity and crack state from the three-dimensional tomographic cloud map, verifies them with the safe vibration velocity threshold, adjusts the blasting parameters in reverse and recalculates in a closed loop, and outputs the optimal parameter configuration.

[0090] This embodiment also provides a computer device applicable to the pre-splitting blasting damage analysis method based on an eccentrically decoupled charge structure, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the pre-splitting blasting damage analysis method based on an eccentrically decoupled charge structure as proposed in the above embodiment.

[0091] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0092] This embodiment also provides a storage medium storing a computer program. When executed by a processor, the program implements the pre-splitting blast damage analysis method based on an eccentrically decoupled charge structure as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0093] In summary, this invention achieves high-fidelity visualization of hidden damage states within the rock mass by: constructing a dynamic finite element numerical model for rock blasting and performing virtual detection inversion using the acoustic travel time principle to obtain theoretical and measured samples characterizing damage features; utilizing generative adversarial networks to fuse and reconstruct features from multiple sources, outputting an asymmetric three-dimensional tomographic cloud map of the damage field with confidence evaluation indicators; and constructing a multi-objective fitness function balancing rock-breaking volume and vibration intensity using a safe vibration velocity threshold as the performance boundary, and implementing variable updates and closed-loop recalculation of blasting design parameters through an optimization algorithm. This results in high-fidelity visualization of hidden damage states within the rock mass, effectively eliminating the cumulative errors and detection blind spots of single theoretical calculations, and improving the safety and rock-breaking efficiency of blasting operations in complex environments.

[0094] 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 it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A pre-splitting blasting damage analysis method based on an eccentrically decoupled charge structure, characterized in that: This includes obtaining the physical and mechanical parameters of the rock mass and the initial blasting design parameters, constructing a numerical model of the rock mass, and setting a safe vibration velocity threshold; Based on the eccentric uncoupled charge geometric model constructed from the initial blasting design parameters, the asymmetric initial borehole wall pressure is calculated as the initial load, non-uniform borehole wall boundary loading is performed, and forward calculation is performed based on the initial load to obtain the theoretical damage field of the rock mass. Virtual probe waves are emitted within a small time difference to perform inversion, obtain the real damage field of the rock mass, and dynamically correct the elastic modulus of the rock mass in real time. The real damage field is a physical damage distribution reconstructed from the reflected signal of the virtual probe wave, which is used to drive the dynamic correction of the elastic modulus of the rock mass in real time. Extract the differential delay time between adjacent boreholes from the initial blasting design parameters, set the differential delay time as the virtual detection time window in the numerical simulation, and pause the application of new explosive loads; Within the virtual detection time window, a virtual broadband detection wave is generated at the preset observation node in the rock mass numerical model to penetrate the rock mass area that has been affected by the previous blasting. Extract the transmitted and reflected wave signals of the probe wave after it passes through the damaged rock mass, and quantitatively calculate the wave velocity reduction rate and amplitude attenuation coefficient. By combining the full waveform inversion algorithm, the wave velocity reduction rate and amplitude attenuation coefficient are mapped and transformed into real damage variables with spatial grid distribution, thereby obtaining the real damage field of the rock mass; Based on the actual damage field, the elastic modulus of each grid cell in the rock mass numerical model is covered and dynamically corrected one by one using the elastic modulus degradation equation, and the corrected rock mass numerical model is used as the loading medium for the next micro-delay detonation. The theoretical damage field is fused with the real damage field to reconstruct the asymmetric damage field and output a three-dimensional tomographic cloud map containing confidence level. Extract the boundary peak vibration velocity and fracture state from the three-dimensional tomographic cloud map, verify them with the safe vibration velocity threshold, adjust the blasting parameters in reverse and recalculate in closed loop, and output the optimal parameter configuration. The three-dimensional tomographic cloud map is a visualized three-dimensional data field generated by jointly visually encoding the reconstructed damage variables and data inversion reliability based on the spatial coordinates of the rock mass.

2. The pre-splitting blasting damage analysis method based on an eccentrically decoupled charge structure as described in claim 1, characterized in that: The physical and mechanical parameters of the rock mass include density, compressive strength, tensile strength, elastic modulus, and initial longitudinal wave velocity; the initial blasting design parameters include borehole diameter, explosive charge diameter, single-hole charge quantity, decoupling coefficient, and inter-hole differential delay time.

3. The pre-splitting blasting damage analysis method based on an eccentrically decoupled charge structure as described in claim 2, characterized in that: The non-uniform borehole wall boundary loading includes establishing a polar coordinate system with the borehole center as the pole and defining the central angle; Based on the aforementioned eccentric uncoupled charge geometric model, the radial expansion distance of the explosive gas from the surface of the charge cartridge to the inner wall of the borehole is calculated at different central angles. Based on the radial expansion distance, the inner wall of the borehole is divided into a wall-attached contact area and a cavity expansion area; The wall-attached contact area is the area where the outer surface of the explosive charge is in direct contact with the inner wall of the borehole, and the explosive shock wave is directly coupled without being attenuated by the gap medium. The initial peak pressure generated by the explosive detonation is directly used as the borehole pressure of the wall-attached contact area. The cavity expansion zone is the area of ​​the borehole wall where there is an initial spatial gap between the outer surface of the explosive cartridge and the inner wall of the borehole, and which is subjected to the effect of the explosive gas after adiabatic expansion and attenuation through the gap. The radial expansion distance is substituted into the adiabatic expansion equation of the explosive gas to calculate the borehole wall pressure after attenuation by the medium. By combining the calculation results of the contact zone with the cavity expansion zone, an asymmetric initial borehole wall pressure distribution function that continuously varies with the central angle is constructed. The asymmetric initial borehole wall pressure distribution function is then mapped and transformed into the mechanical boundary conditions of the borehole wall grid nodes in the rock mass numerical model, thus completing the non-uniform loading.

4. The pre-splitting blasting damage analysis method based on an eccentrically decoupled charge structure as described in claim 3, characterized in that: The reconstructed asymmetric damage field includes constructing a generative adversarial network model that includes a generator network and a discriminator network; The theoretical damage field, the actual damage field, and the spatial coordinates of the grid cells in the rock mass numerical model are extracted and spliced ​​into tensors. The spliced ​​tensors are then input into the generator network to output the predicted damage field. The theoretical damage field and the actual damage field are used as theoretical samples and measured samples, respectively, and input with the predicted damage field into the discriminator network to calculate and output the discrimination probability of the predicted damage field; A composite loss function is constructed, which includes adversarial loss, data fidelity loss, and physical mechanism loss. The adversarial loss is the deviation between the discrimination probability and a preset threshold. The data fidelity loss is the calculation error between the predicted damage field and the real damage field at the virtual detection node. The physical mechanism loss is the calculation residual obtained by substituting the spatial gradient of the predicted damage field into the rock mass wave equation. The generator network is trained alternately using the composite loss function until convergence, and the output of the generator network after convergence is used as the asymmetric damage field. The measured data obtained after the explosion are updated to update the measured samples, and the generative adversarial network model is incrementally trained.

5. The pre-splitting blasting damage analysis method based on an eccentrically decoupled charge structure as described in claim 4, characterized in that: The output, a three-dimensional tomographic cloud map containing confidence, includes obtaining the spatial coordinates and damage variables of each grid cell in the asymmetric damage field, and simultaneously extracting the data confidence parameters corresponding to each grid cell. Construct a multidimensional visual mapping rule to map the damage variable to a first visual attribute and the data confidence parameter to a second visual attribute; Perform three-dimensional spatial interpolation based on the spatial coordinates of each grid cell, the first visual attribute, and the second visual attribute; The interpolated spatial data is reconstructed using a 3D volume rendering algorithm to output the 3D tomographic cloud map containing confidence level.

6. The pre-splitting blasting damage analysis method based on an eccentrically decoupled charge structure as described in claim 5, characterized in that: The reverse adjustment of blasting parameters and closed-loop recalculation includes obtaining the peak vibration velocity distribution data of the three-dimensional tomographic cloud map at the preset protection boundary, as well as the effective damage volume within the target blasting area. A preset target vibration velocity distribution is introduced, and the distribution distance measure between the peak vibration velocity distribution data and the target vibration velocity distribution is calculated. A multi-objective fitness function F is constructed with the goal of minimizing the distribution distance measure and maximizing the effective damage volume. The hard constraint condition for the multi-objective fitness function F is set as the peak velocity distribution data. The maximum value in the range is less than or equal to the safe vibration velocity threshold. ; Based on the fitness function F and the hard constraints, the initial blasting design parameter set is updated using an optimization algorithm to obtain the new blasting parameter set for the current iteration. The new blasting parameter set is imported into the rock mass numerical model and the steps of forward blasting loading, virtual probe wave inversion and asymmetric damage field reconstruction are re-executed to generate an updated three-dimensional tomographic cloud map. The variable update and reconstruction steps are executed repeatedly until the fitness function converges and satisfies the hard constraint condition. The new burst parameter set in the current convergent state is then extracted as the optimal parameter configuration.

7. A pre-splitting blasting damage analysis system based on an eccentrically decoupled charge structure, based on the pre-splitting blasting damage analysis method based on an eccentrically decoupled charge structure as described in any one of claims 1 to 6, characterized in that: include, The data acquisition unit obtains the physical and mechanical parameters of the rock mass and the initial blasting design parameters, constructs a numerical model of the rock mass, and sets a safe vibration velocity threshold. The deduction unit, based on the eccentric uncoupled charge geometric model constructed from the initial blasting design parameters, calculates the asymmetric initial borehole wall pressure as the initial load, performs non-uniform borehole wall boundary loading, and performs forward calculation based on the initial load to obtain the theoretical damage field of the rock mass. The inversion unit emits virtual probe waves within a micro-difference delay time to perform inversion, obtain the real damage field of the rock mass, and dynamically correct the elastic modulus of the rock mass in real time. The fusion unit fuses the theoretical damage field with the real damage field to reconstruct the asymmetric damage field and outputs a three-dimensional tomographic cloud map containing confidence level. The optimization unit extracts the boundary peak vibration velocity and crack state from the three-dimensional tomographic cloud map, verifies them with the safe vibration velocity threshold, adjusts the blasting parameters in reverse and recalculates in a closed loop, and outputs the optimal parameter configuration.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the pre-splitting blasting damage analysis method based on an eccentric uncoupled charge structure as described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the pre-splitting blasting damage analysis method based on the eccentric uncoupled charge structure as described in any one of claims 1 to 6.

Images