Golden mine working face self-adaptive mining system based on digital twinning

Abstract

The invention relates to the technical field of digital twinning intelligent mining, in particular to a digital twinning-based gold mine working face self-adaptive mining system. And the geological modeling module constructs a composite geological twinborn body fusing a rock mass damage quantification field, an ore body topological structure and a weak plane permeability tensor. And the mining optimization module performs closed-loop calculation based on the twinborn body, and outputs a mining parameter dynamic scheme and a geological risk zoning map. And the aging correction module performs long-term mechanical feedback correction on the mining scheme according to the matching relationship between the rock mass rheological property and the stress field evolution trajectory. And the collaborative intervention module generates a regulation and control instruction according to the risk map. According to the system, self-adaptive optimization of the mining process on the dynamic change of the geological environment and advanced accurate intervention of engineering risks are realized.

Description

Technical Field

[0001] This invention relates to the field of digital twin intelligent mining technology, and in particular to an adaptive mining system for gold mine working faces based on digital twins. Background Technology

[0002] Currently, in the field of underground metal mining, the application of digital twin technology is mostly concentrated on the virtual mapping of production equipment and processes, or the establishment of static, geometric three-dimensional geological models. These models mainly reflect basic attributes such as ore body morphology and fixed grade distribution, and the geological information is fragmented and static. There is a lack of effective quantitative characterization and integrated methods for the evolution of rock mass damage induced by mining activities, the topological relationships of complex ore body spatial structures, and the directional control of seepage fields by tectonic weak surfaces. This makes it difficult for digital models to accurately depict the dynamic geological response of the multi-field coupling of "stress-structure-seepage" under mining, limiting the depth of precise decision-making on mining processes based on digital twins.

[0003] The optimization and adjustment of existing mining schemes largely rely on pre-mining exploration data and short-term on-site monitoring feedback, resulting in a lag in the decision-making process. Traditional methods for considering rock mass mechanical behavior are mostly based on instantaneous or elastoplastic analysis, generally neglecting the time effect of rock mass rheological properties, i.e., the gradual deterioration of rock mass strength and deformation over time. Because the dynamic evolution trajectory of the stress field caused by mining is not systematically correlated with the time-dependent damage mechanism of the rock mass, it is impossible to make long-term, predictive adjustments to key processes such as mining sequence, support parameters, and advance speed. The setting of mining parameters often becomes mismatched with the actual geological environment, where creep damage has already occurred, in the later stages of the cycle, increasing engineering risks such as local instability and water inrush. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and propose an adaptive mining system for gold mine working faces based on digital twins.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: an adaptive mining system for gold mine working faces based on digital twins, comprising: The geological dynamic perception module acquires the full-cycle geological dynamic dataset generated by the mining face during mining activities. The full-cycle geological dynamic dataset includes the evolution trajectory of the ore body's original stress field, the spatial variation curve of grade, and the response data of the surrounding rock structure. The geological modeling module drives the construction of a hierarchically coupled digital twin geological body through the full-cycle geological dynamic dataset. The digital twin geological body is expressed as a composite geological characterization that includes a rock mass damage quantification field, a continuous topological structure of the ore body, and a permeability tensor of tectonic weak surfaces. The mining optimization module performs closed-loop optimization calculations on mining behavior and geological conditions based on the composite geological characterization, and extracts dynamic configuration schemes for mining process parameters and geological disturbance risk zoning maps during the working face advancement process. The time-based correction module performs long-term mechanical feedback correction on the dynamic configuration scheme of the mining process parameters based on the time-based damage mechanism of the rock mass, and outputs the time-based corrected dynamic configuration scheme of the mining process parameters. The long-term mechanical feedback correction is related to the matching relationship between the evolution trajectory of the original stress field of the ore body and the rheological properties of the rock mass. The collaborative intervention module uses the geological disturbance risk zoning map to derive a set of collaborative intervention instructions for the mining process.

[0006] Preferably, the construction of a hierarchically coupled digital twin geological body driven by the full-cycle geological dynamic dataset includes: The evolution trajectory of the original stress field of the ore body is deconstructed into multiple stress state segments according to the mining time sequence, and each stress state segment corresponds to a specific mining cycle. For each stress state segment, the following steps are performed: The surrounding rock structure response data is integrated to generate a multi-dimensional deformation collaborative network of the working face surrounding rock system. The multi-dimensional deformation collaborative network contains a roof delamination distribution cloud map, a two-sided convergence rate field and a floor activation intensity spectrum. The multidimensional deformation collaborative network is subjected to deep spatiotemporal correlation analysis with the current stress state segment to form a geological state snapshot corresponding to the current mining cycle. The geological state snapshot encapsulates the rock mass damage evolution tensor, the geometric description of the ore body mining boundary, and the water-conducting fracture development trend matrix. A progressive overlay evolutionary deduction was performed on multiple geological state snapshots arranged in chronological order to calculate the quantitative field of rock mass damage, the continuous topological structure of the ore body, and the permeability tensor of tectonic weak surfaces; among which... The rock mass damage quantification field is defined as the result of the convolution operation of the cumulative damage gradient along the mining direction of the rock mass damage evolution tensor and the energy dissipation rate of the deformation cooperative network. The continuous topological structure of the ore body is expressed as a combination of connectivity measures and morphological persistence features in three-dimensional space for the geometric description of the ore body's mining boundary. The permeability tensor of the structural weak surface is characterized as the anisotropic distribution of permeability within a unit geological volume element, representing the development trend matrix of water-conducting fractures.

[0007] Preferably, the step of performing deep spatiotemporal correlation analysis between the multi-dimensional deformation collaborative network and the current stress state segment to form a geological state snapshot corresponding to the current mining cycle includes: Based on the dependence between the delamination distribution cloud map of the top plate and the principal stress direction in the current stress state segment, a delamination principal stress coupled response model is established. The spatial distribution function of the rock mass damage evolution tensor is obtained by iteratively solving the delamination principal stress coupled response model. Based on the correlation mapping between the convergence rate field of the two sides and the tectonic stress concentration coefficient, a dynamic evaluation framework for the stability of the two sides is constructed. The dynamic evaluation framework for the stability of the two sides incorporates the real-time attenuation parameters of the rock mass strength softening coefficient and the shear strength of the structural surface. Combining the activation intensity spectrum of the base plate with the triggering conditions of water-rock interaction, a multi-step judgment logic for the risk of water inrush on the base plate is formed. The multi-step judgment logic incorporates a bidirectional feedback mechanism between pore water pressure propagation and fracture propagation rate. The spatial distribution function, the output of the dynamic evaluation framework for stability of the two sides, and the conclusion of the multi-step judgment logic for the risk of water inrush in the bottom plate are assimilated and fused into three-dimensional field data to generate a high-fidelity geological field data set that integrates damage state, ore body geometry and hydraulic characteristics.

[0008] Preferably, the step of performing closed-loop optimization calculations based on the composite geological characterization to determine the mining behavior and geological state, and extracting dynamic configuration schemes for mining process parameters and geological disturbance risk zoning maps during the working face advancement process, includes: The quantitative field of rock mass damage is imported into the damage assessment stage of the pre-set intelligent decision kernel, and the spatial coordinate set of potential collapse zone and critical instability stress envelope surface are identified by the instability state projection module. The continuous topology of the ore body is imported into the resource optimization mining stage of the intelligent decision kernel, and a mining process adaptability assessment is performed to generate a mining continuity index and a predicted spectrum of ore lean ratio. The permeability tensor of the constructed weak surface is imported into the water hazard early warning link of the intelligent decision kernel, and the evolution direction of the dominant water-conducting channel and the inrush risk intensity value are calculated based on the seepage mutation identification model. By integrating the critical instability stress envelope, the ore lean loss ratio prediction spectrum, and the water inrush risk intensity value, a comprehensive disturbance index for the working face is constructed. Based on the comparison between the comprehensive disturbance index of the working face and the dynamic safety boundary, a dynamic configuration scheme for mining process parameters is determined. Based on the geometric spatial clustering results of the spatial coordinate set, the mining continuity index, and the evolution direction of the dominant water-diverting channel, the spatial boundaries of the landslide risk zone, the resource inefficiency zone, and the water hazard threat zone are drawn.

[0009] Preferably, the step of performing long-term mechanical feedback correction on the dynamic configuration scheme of mining process parameters based on the time-dependent damage mechanism of rock mass, and outputting the time-dependent corrected dynamic configuration scheme of mining process parameters, includes: The dominant stress evolution path and stress relaxation time constant in the evolution trajectory of the original stress field of the ore body are analyzed, and the time-varying reduction of the long-term strength parameters of the rock mass due to mining disturbance is calculated. Based on the time-varying reduction, stress history compensation calculation is performed on the rock mass damage quantification field to generate a time-varying corrected rock mass damage quantification field. By associating the stress relaxation time constant with the constitutive relationship of rock mass creep damage, creep fracture correction deduction is performed on the continuous topology of the ore body to generate a time-varying corrected ore body continuous topology. Based on the cumulative data of rock mass fatigue damage under the dominant stress path, the permeability tensor of the structural weak surface is subjected to fatigue damage coupling adjustment processing to generate a time-varying corrected permeability tensor of the structural weak surface. The time-varying modified rock mass damage quantification field, ore body continuity topology, and tectonic weak surface permeability tensor are re-input into the intelligent decision kernel for iterative calculation, and the dynamic configuration scheme of mining process parameters after long-term mechanical feedback correction is output.

[0010] Preferably, the step of performing stress history compensation calculation on the rock mass damage quantification field based on the time-varying reduction to generate a time-varying corrected rock mass damage quantification field includes: The design time-history strength degradation function is constructed by calling the rock mass strength benchmark parameter set of the working face under the original geostress environment and the time-varying reduction. The cumulative damage factor of the rock mass is calculated using the design time-history strength degradation function. The cumulative damage factor of the rock mass is characterized as the product of the stress path integral and the time-varying rate of the strength parameter. The cumulative damage factor of the rock mass is injected into the generation process of the rock mass damage quantification field to form a correction amount of the rock mass damage quantification field containing the historical damage accumulation effect. The rock mass damage quantification field correction is applied to compensate for the rock mass damage self-healing effect, which depends on the synergistic effect of fracture closure rate and time healing factor.

[0011] Preferably, the method of deriving a collaborative intervention instruction set for the mining process using the geological disturbance risk zoning map includes: Facing the spatial boundary of the collapse risk zone, the optimal roof collaborative control trajectory is calculated. The optimal roof collaborative control trajectory is achieved by redistributing the support strength and step distance of the hydraulic support group. To address the spatial boundaries of the resource-inefficient zone, a replanning scheme for the mining trajectory is designed. This scheme includes curvature optimization of the adaptive cutting path of the coal mining machine and load matching adjustment of the cutting power. Based on the spatial boundaries of the water hazard threat zone, a dynamic grouting intervention strategy is formulated. The dynamic grouting intervention strategy dynamically matches the grouting timing and grout diffusion radius based on the changing trend of the water inrush risk intensity value. The optimal roof collaborative control trajectory, the mining trajectory replanning scheme, and the dynamic grouting intervention strategy are optimized and sorted by task scheduling to form a set of collaborative intervention instructions containing execution sequences and control thresholds; The set of collaborative intervention instructions for the mining process is specified as a strategy for adjusting the intensity of roof maintenance and a scheme for reorganizing the mining sequence.

[0012] Preferably, the design forms a replanning scheme for the mining trajectory, including: Extract the spatial morphological features of the resource inefficiency zone and calculate the effective thickness variation function and strike variation rate of the ore body; The rotation speed control strategy of the coal mining machine drum is dynamically adapted based on the effective thickness variation function of the ore body, and the rotation speed control strategy is synchronized with the local extreme value change trend of the effective thickness. The stability control law of the coal mining machine traction mechanism is optimized in real time based on the aforementioned change rate of trajectory, so that the fluctuation amplitude of the traction acceleration is constrained to be proportional to the reciprocal of the change rate of trajectory. The cutting depth of the cutting teeth is dynamically configured based on the hardness identification results of the rock-ore interface to ensure that the cutting ratio energy consumption is within the preset economic operating range. A mapping table for adjusting mining process parameters is generated, which integrates rotation speed control strategy, stability control law parameters, and cutting depth configuration.

[0013] Preferably, the calculation of the evolution direction of the dominant water-conducting channel and the inrush risk intensity value based on the seepage mutation identification model includes: Obtain the data sequence of the permeability tensor of the constructed weak surface under multiple consecutive mining time series; Calculate the permeability variation gradient field of each geological element between adjacent time series based on the data sequence; Identify connected regions in the permeability change gradient field where the gradient value exceeds a preset threshold, and mark each connected region as a potential water-conducting channel; Calculate the spatial displacement vector of the geometric center point of each potential water-conducting channel between the current time series and the previous time series, and determine the direction of the spatial displacement vector as the current evolution direction of the potential water-conducting channel; The maximum gradient value of the permeability change gradient field in each potential water-conducting channel is extracted, and combined with the volume of the potential water-conducting channel, the initial value of the inrush risk intensity of each potential water-conducting channel is obtained through weighted fusion calculation. Based on the rock mechanics properties reflected by the permeability tensor of the structural weak surface, the initial value of the inrush risk intensity of each potential water-conducting channel is corrected for rock stability, and the inrush risk intensity value of each potential water-conducting channel is output.

[0014] Preferably, the system further includes: Within the set model verification time window, the measured roof pressure step distance data and actual ore recovery rate data of the mining face are collected synchronously. The measured data of the pressure step distance of the roof is compared with the pressure step distance distribution predicted by the digital twin geological body to generate a pressure step distance prediction deviation correction factor. The actual ore recovery rate data and the predicted recovery rate derived from the composite geological characterization are evaluated for regional consistency, and a recovery rate prediction deviation correction factor is generated. The evaluation logic of the intelligent decision kernel is adaptively calibrated using the step distance prediction deviation correction factor and the recovery rate prediction deviation correction factor to generate a calibrated intelligent decision kernel. The calibrated intelligent decision kernel will be deployed and applied to the adaptive decision-making process for mining in future working faces.

[0015] Compared with the prior art, the advantages and positive effects of the present invention are as follows: A composite geological digital twin was constructed, integrating a quantitative field of rock mass damage, the continuous topological structure of the ore body, and the permeability tensor of tectonic weak surfaces. This model integrates the spatial quantitative distribution of fracture development within the rock mass during mining, the topological connectivity description of the complex ore body morphology, and the tensor-based permeability characteristics of geological weak surfaces as directional seepage channels into a unified, evolving twin model. This achieves an integrated and refined digital mapping of the coupled evolution of the mechanical damage state of the geological body, resource endowment structure, and potential seepage paths under underground engineering disturbances. The geological state information used for mining optimization calculations has shifted from traditional static geometry and attributes to a composite field encompassing dynamic damage propagation, structural connectivity networks, and dominant seepage directions, enhancing the completeness of the geological realism and scientific basis for mining behavior simulation and process parameter optimization.

[0016] A long-term mechanical feedback correction mechanism based on the matching relationship between the time-dependent damage mechanism of rock mass and the evolution trajectory of stress field is established. This scheme continuously analyzes the dynamic evolution data of the original stress field of the ore body acquired by the sensing module through a time-dependent correction module, and performs real-time matching and coupling calculations with the inherent rheological properties of the rock mass. Based on this time-related mechanical matching relationship, the initial process parameter configuration scheme output by the mining optimization module is proactively corrected through mechanical feedback. This allows the formulation of mining parameters to not only respond to the current geological conditions but also proactively anticipate changes in the geomechanical environment caused by stress relaxation and creep damage accumulation over a future period. This enables the mining process parameters to automatically and dynamically adapt to the time-varying characteristics of rock mass mechanical behavior as the mining face advances and time progresses, achieving a leap from "instantaneous static adaptation" to "full-cycle dynamic compliance," reducing long-term engineering risks caused by the mismatch between parameters and the time-varying geological environment. Attached Figure Description

[0017] Figure 1 This is a timing diagram of the adaptive mining system for gold mine working faces based on digital twins as described in this invention. Figure 2 Flowchart for constructing a digital twin geological body; Figure 3 A flowchart for closed-loop optimization calculation and risk partitioning; Figure 4 A dual-axis dynamic monitoring diagram for roof coordination control during the mining process of a gold mine working face; Figure 5 This is a cloud map showing the spatial distribution of gold ore grades. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0019] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0020] The overall implementation scheme of the adaptive mining system for gold mine working faces based on digital twins is as follows: The system acquires a full-cycle geological dynamic dataset generated during mining activities through a geological dynamic perception module. This dataset covers the evolution trajectory of the ore body's original stress field, grade spatial variation curves, and surrounding rock structural response data. The geological modeling module then drives the construction of a layered, coupled digital twin geological body using the full-cycle geological dynamic dataset. This geological body is expressed as a composite geological characterization including a quantitative field of rock mass damage, the continuous topological structure of the ore body, and the permeability tensor of tectonic weak surfaces. The mining optimization module performs closed-loop optimization calculations based on this composite geological characterization, extracting dynamic configuration schemes for mining process parameters and geological disturbance risk zoning maps during the working face advancement process. The time-based correction module performs long-term mechanical feedback correction on the dynamic configuration schemes of mining process parameters based on the time-based damage mechanism of the rock mass, outputting the time-corrected dynamic configuration schemes. This correction process is related to the matching relationship between the evolution trajectory of the ore body's original stress field and the rheological properties of the rock mass. The collaborative intervention module derives a collaborative intervention instruction set for the mining process using the geological disturbance risk zoning map.

[0021] In one embodiment of the present invention, see [reference] Figure 2 In specific implementation, the adaptive mining system for gold mine working faces based on digital twins constructs a layered coupled digital twin geological body. The implementation method includes deconstructing the evolution trajectory of the original stress field of the ore body into multiple stress state segments according to the mining time sequence. Each stress state segment corresponds to a specific mining cycle. For each stress state segment, the surrounding rock structure response data is integrated to generate a multi-dimensional deformation collaborative network of the working face surrounding rock system. The multi-dimensional deformation collaborative network contains a cloud map of the roof delamination distribution, the convergence rate field of the two sides, and the activation intensity spectrum of the floor. The multi-dimensional deformation collaborative network is subjected to in-depth spatiotemporal correlation analysis with the current stress state segment to form a geological state snapshot corresponding to the current mining cycle. The geological state snapshot encapsulates the rock mass damage evolution tensor, the geometric description of the ore body mining boundary, and the water-conducting fracture development trend matrix. The multiple geological state snapshots arranged in time sequence are subjected to progressive superposition evolution and deduction to calculate the rock mass damage quantification field, the ore body continuity topology, and the permeability tensor of the structural weak surface.

[0022] In some embodiments, deep spatiotemporal correlation analysis establishes a delamination principal stress coupling response model based on the dependence of the distribution cloud map of the top plate delamination and the principal stress direction in the current stress state segment. The spatial distribution function of the rock mass damage evolution tensor is obtained by iteratively solving the delamination principal stress coupling response model. A dynamic evaluation framework for the stability of the two sides is constructed based on the correlation mapping between the convergence rate field of the two sides and the tectonic stress concentration coefficient. The dynamic evaluation framework for the stability of the two sides incorporates the real-time attenuation parameters of the rock mass strength softening coefficient and the shear strength of the structural surface. A multi-step judgment logic for the risk of water inrush in the bottom plate is formed by combining the activation intensity spectrum of the bottom plate and the triggering conditions of water-rock interaction. The multi-step judgment logic for the risk of water inrush in the bottom plate embeds a bidirectional feedback mechanism between pore water pressure propagation and fracture propagation rate. The spatial distribution function, the output of the dynamic evaluation framework for the stability of the two sides, and the conclusion of the multi-step judgment logic for the risk of water inrush in the bottom plate are assimilated and fused in three-dimensional field data to generate a high-fidelity geological field data set that integrates damage state, ore body geometry, and hydraulic characteristics.

[0023] In practical implementation, the rock mass damage quantification field is defined as the result of the convolution operation of the cumulative damage gradient along the mining direction of the rock mass damage evolution tensor and the energy dissipation rate of the deformation co-network. The ore body continuity topology is expressed as a combination of the connectivity measure and morphological persistence characteristics of the ore body mining boundary geometry in three-dimensional space. The permeability tensor of the structural weak surface is characterized as the anisotropic distribution of the water-conducting fracture development status matrix with respect to permeability within a unit geological volume element. It can be understood that the numerical value of the rock mass damage quantification field is calculated through a dimensionless quantization function that integrates the spatial correlation between the damage gradient and the energy dissipation rate. Its mathematical expression is:

[0024] in: Represents spatial coordinates The quantized field value of rock mass damage at a given location is a dimensionless scalar. Represents the rock mass damage evolution tensor In position The norm of the gradient tensor at that point, This indicates that the preset gradient baseline value is used for normalization. Indicates the location of the deformable cooperative network. The scalar value of the energy dissipation rate at that location. This indicates that a preset energy dissipation rate benchmark value is used for normalization. Optionally, the progressive superposition evolutionary deduction is implemented through a recurrent neural network model with memory units. This model takes a time-series of geological state snapshots as input sequences, recursively updates the internal states of the rock mass damage quantification field, the ore body continuity topology, and the permeability tensor of the tectonic weak surface, and outputs the final calculation results.

[0025] In some embodiments, the generation of the multidimensional deformation collaborative network relies on a microseismic monitoring array, a distributed fiber optic strain sensing system, and a borehole camera deployed within the surrounding rock of the working face. The roof delamination distribution cloud map is obtained by joint inversion of spatial clustering and energy release rate of microseismic events. The convergence rate field of the two sides is calculated from the distributed fiber optic strain sensing data through a strain-displacement transformation model. The floor activation intensity spectrum is obtained by fusing and analyzing the fracture aperture change sequence observed by the borehole camera with stress monitoring data. It can be understood that the connectivity measure in the continuous topology of the ore body is obtained by calculating the graph theory characteristic parameters of the network formed by the geometric description of the ore body mining boundary in three-dimensional space, and the morphological persistence characteristics are obtained by analyzing the topological invariance of the boundary geometry over time. Optionally, the principal direction and magnitude of the permeability tensor of the structural weak surface are determined by eigenvalue decomposition of the water-conducting fracture development trend matrix, and the anisotropic distribution is described by the ratio of eigenvalues.

[0026] In one embodiment of the present invention, see [reference] Figure 3 In practical implementation, the adaptive mining system for gold mine working faces based on digital twins performs closed-loop optimization calculations on mining behavior and geological conditions, extracting dynamic configuration schemes for mining process parameters and geological disturbance risk zoning maps during the working face advancement process. Specifically, the quantitative field of rock mass damage is imported into the damage assessment stage of the pre-set intelligent decision kernel, using the instability state projection module to identify the spatial coordinate set of potential collapse zones and the critical instability stress envelope. The ore body continuity topology is imported into the resource optimization stage of the intelligent decision kernel, performing a mining process adaptability assessment and generating a mining continuity index and ore lean-loss ratio prediction spectrum. The permeability tensor of structural weak surfaces is imported into the water hazard early warning stage of the intelligent decision kernel, calculating the evolution direction of dominant water-conducting channels and the water inrush risk intensity value based on the seepage mutation identification model. In essence, by integrating the critical instability stress envelope, the ore lean-loss ratio prediction spectrum, and the water inrush risk intensity value, a comprehensive working face disturbance index is constructed, and the dynamic configuration scheme for mining process parameters is determined based on the comparison between the comprehensive working face disturbance index and the dynamic safety boundary. In practice, based on the geometric spatial clustering results of the spatial coordinate set, the mining continuity index and the evolution direction of the dominant water diversion channel, the spatial boundaries of the landslide risk zone, the resource inefficiency zone and the water hazard threat zone are drawn. The geological disturbance risk zoning map is composed of the spatial boundaries of the landslide risk zone, the resource inefficiency zone and the water hazard threat zone.

[0027] In some embodiments, the damage assessment stage of the intelligent decision-making kernel includes a numerical simulation engine coupled with finite element and discrete element methods. The instability state projection module calculates the stability coefficient of each point in the rock mass damage quantification field, gathers all spatial points with stability coefficients below a preset threshold into a spatial coordinate set, and fits the critical instability stress envelope surface surrounding these points. In a specific implementation, the process of the seepage mutation identification model calculating the evolution direction of the dominant water-conducting channel and the water inrush risk intensity value includes acquiring the data sequence of the permeability tensor of the structural weak surface under multiple consecutive mining time sequences, calculating the permeability change gradient field of each geological element between adjacent time sequences based on the data sequence, identifying connected regions in the permeability change gradient field where the gradient value exceeds a preset threshold, and marking each connected region as a potential water-conducting channel. The spatial displacement vector of the geometric center point of each potential water-conducting channel between the current time sequence and the previous time sequence is calculated, and the direction of the spatial displacement vector is determined as the current evolution direction of the potential water-conducting channel. The maximum gradient value of the permeability variation gradient field within each potential water-conducting channel is extracted, and combined with the volume of the potential water-conducting channel, a weighted fusion calculation is performed to obtain the initial value of the inrush risk intensity for each potential water-conducting channel. Optionally, the permeability variation gradient field... The calculation uses the central difference method, and its expression is: in: Indicates the location in grid coordinates Geological elements in time Relative to time The gradient vector of permeability change, Indicates time At the time of location The principal permeability scalar value of the permeability tensor of the weak surface is constructed. Indicates time At the time of location The principal permeability scalar value of the permeability tensor of the weak surface is constructed. , , These represent the dimensions of the geological volume element in three spatial directions. Based on the surrounding rock mechanical properties reflected by the permeability tensor of the tectonic weak surface, the initial value of the inrush risk intensity of each potential water-conducting channel is corrected for surrounding rock stability, and the inrush risk intensity value of each potential water-conducting channel is output. In some embodiments, the surrounding rock stability correction is achieved through a reduction coefficient function related to the surrounding rock cohesion and internal friction angle.

[0028] Optional, comprehensive disturbance index of the working face Instability risk value characterized by the critical instability stress envelope surface Resource inefficiency as characterized by the ore lean-loss ratio prediction spectrum and the intensity value of water inrush risk We get the weighted sum, that is ,in , , These are weighting coefficients set based on the principles of mining safety and economic benefits. It can be understood that the dynamic safety boundary is a threshold surface that is updated according to changes in mining progress and geological conditions. When the comprehensive disturbance index of the working face... When the value in a certain area exceeds the dynamic safety boundary, the dynamic configuration scheme of mining process parameters will generate parameter adjustment instructions for that area, such as reducing the mining height or adjusting the advance speed.

[0029] In one embodiment of the present invention, in a specific implementation, the adaptive mining system for a gold mine working face based on digital twins performs long-term mechanical feedback correction on the dynamic configuration scheme of mining process parameters according to the time-dependent damage mechanism of the rock mass, and outputs the time-dependent corrected dynamic configuration scheme of mining process parameters. The implementation method includes analyzing the dominant stress evolution path and stress relaxation time constant in the evolution trajectory of the original stress field of the ore body, calculating the time-varying reduction of the long-term strength parameters of the rock mass due to mining disturbance, performing stress history compensation calculation on the quantitative field of rock mass damage based on the time-varying reduction, and generating the time-varying corrected quantitative field of rock mass damage. By correlating the stress relaxation time constant with the constitutive relationship of rock mass creep damage, creep fracture correction is performed on the continuous topology of the ore body to generate a time-varying corrected ore body continuous topology. Based on the cumulative data of rock mass fatigue damage under the dominant stress path, fatigue damage coupling adjustment is performed on the permeability tensor of the structural weak surface to generate a time-varying corrected structural weak surface permeability tensor. The time-varying corrected rock mass damage quantification field, the continuous topology of the ore body, and the permeability tensor of the structural weak surface are re-input into the intelligent decision kernel for iterative calculation, and a dynamic configuration scheme of mining process parameters after long-term mechanical feedback correction is output.

[0030] In practice, the process of performing stress history compensation calculation on the rock mass damage quantification field based on the time-varying reduction includes calling the rock mass strength benchmark parameter set and time-varying reduction of the working face under the original geostress environment, constructing the design time-history strength degradation function, calculating the rock mass cumulative damage factor through the design time-history strength degradation function, which is characterized as the product of stress path integral and strength parameter time-varying rate, injecting the rock mass cumulative damage factor into the generation process of the rock mass damage quantification field to form a rock mass damage quantification field correction quantity containing the historical damage accumulation effect, and applying rock mass damage self-healing effect compensation to the rock mass damage quantification field correction quantity. The rock mass damage self-healing effect compensation depends on the synergistic effect of fracture closure rate and time healing factor.

[0031] In some embodiments, design time history intensity degradation function It is expressed as the ratio of the long-term strength parameter of the rock mass to the initial strength reference parameter, and is a value that varies with time. Decreasing dimensionless function, cumulative damage factor of rock mass The calculation is performed using the normalized integral of the dominant stress evolution path and the strength degradation rate, with the specific formula as follows:

[0032] in: The cumulative damage factor of the rock mass is a dimensionless scalar. This indicates the total time span corresponding to the current mining activity. Indicates time The dominant stress evolution path vector, This indicates the magnitude of the dominant stress at that moment. This represents the benchmark parameters of rock mass strength at the working face under the original geostress environment. The design time history intensity degradation function represents the time... The value, This represents the time derivative of the intensity degradation function. This represents the time-varying element. It can be understood that the quantitative field correction for rock mass damage is obtained by applying the cumulative damage factor of the rock mass. As a multiplicative correction coefficient, it is applied to the original quantitative field value of rock mass damage. Optionally, the compensation for the self-healing effect of rock mass damage is achieved through a decay function proportional to the time healing factor, which is determined based on experimental data of rock mass mineral composition and underground environmental humidity.

[0033] In some embodiments, the process of creep fracture correction simulation of the orebody continuity topology employs a viscoelastic mechanics model, using the stress relaxation time constant as the core parameter of the model. This simulates the creep deformation of the rock mass under long-term stress until micro-fracture occurs, thereby updating the connectivity measures and morphological persistence characteristics in the orebody continuity topology. It can be understood that when performing fatigue damage coupling adjustment on the permeability tensor of the structural weak surface, the amplitude and mean data of each stress cycle under the dominant stress path need to be input. The cumulative fatigue damage data is extracted from the stress evolution path using the rainflow counting method, and the anisotropic distribution of the permeability tensor is readjusted according to the fatigue crack propagation direction. Optionally, re-inputting the time-varying corrected data into the intelligent decision kernel for iterative calculation is an automatic loop process. The intelligent decision kernel uses the same evaluation logic as the one used to generate the initial dynamic configuration scheme of mining process parameters, but the input composite geological characterization data has been time-corrected until the change in the output dynamic configuration scheme of mining process parameters is less than a preset convergence threshold.

[0034] In one embodiment of the present invention, in a specific implementation, the adaptive mining system for gold mine working faces based on digital twins utilizes a geological disturbance risk zoning map to derive a set of collaborative intervention instructions for the mining process. The implementation includes calculating the optimal roof collaborative control trajectory based on the spatial boundary of the collapse risk zone. The optimal roof collaborative control trajectory is achieved by redistributing the support strength and step distance coordination relationship of the hydraulic support group. A mining trajectory replanning scheme is designed for the spatial boundary of the resource inefficient zone. The mining trajectory replanning scheme includes curvature optimization of the adaptive cutting path of the coal mining machine and load matching adjustment of the cutting power. A dynamic grouting intervention strategy is formulated based on the spatial boundary of the water hazard threat zone. The dynamic grouting intervention strategy dynamically matches the grouting timing and slurry diffusion radius based on the changing trend of the water inrush risk intensity value. The optimal roof collaborative control trajectory, the mining trajectory replanning scheme, and the dynamic grouting intervention strategy are optimized and sorted by task scheduling to form a set of collaborative intervention instructions containing execution sequences and control thresholds. The set of collaborative intervention instructions for the mining process is specifically manifested as a roof maintenance intensity adjustment strategy and a mining sequence reorganization scheme.

[0035] In practical implementation, the process of designing and forming a mining trajectory replanning scheme includes extracting the spatial morphological characteristics of resource inefficient areas, calculating the effective thickness variation function and strike variation rate of the ore body, dynamically adapting the rotation speed control strategy of the coal mining machine drum based on the effective thickness variation function of the ore body, synchronizing the rotation speed control strategy with the local extreme value change trend of the effective thickness, optimizing the stability control law of the coal mining machine traction mechanism in real time based on the strike variation rate, so that the fluctuation amplitude of the traction acceleration is proportional to the reciprocal of the strike variation rate, dynamically configuring the cutting depth of the cutting teeth in combination with the hardness identification results of the rock-ore interface, ensuring that the cutting ratio energy consumption is within the preset economic operating range, and generating a mining process parameter adjustment mapping table that integrates the rotation speed control strategy, stability control law parameters, and cutting depth configuration.

[0036] In some embodiments, a distributed optimization algorithm is used to calculate the optimal roof collaborative control trajectory. The algorithm aims to minimize the total energy consumption of the hydraulic support group, and uses the roof pressure distribution defined by the critical instability stress envelope as a constraint condition to solve for the optimal support strength setpoint fi of each hydraulic support and the difference in advance step distance between adjacent supports. The optimal roof collaborative control trajectory is determined by the sequence {( , This can be understood as the amplitude of the traction acceleration fluctuation. With the rate of change of direction The constraint relationship between them is defined by the following formula:

[0037] in: This represents the fluctuation amplitude of the coal mining machine's traction acceleration within one control cycle, and its dimension is [length] / [time]. 2 , This represents the rate of change of the local strike of the working face, calculated from the continuous topology of the ore body, and its dimension is [angle] / [length]. It is a proportionality coefficient related to the mechanical characteristics and preset stability level of the coal mining machine, and its dimension is [length]. 2 / ([time] 2 • [Angle]). Optionally, the hardness identification of the rock-ore interface is achieved through vibration sensors and current sensors installed on the coal mining machine drum, and the hardness identification results are mapped to different cutting depth levels. In some embodiments, the execution logic of the dynamic grouting intervention strategy is to trigger a grouting command when the water inrush risk intensity value of a certain area exceeds the rising threshold for three consecutive monitoring cycles. The grout diffusion radius is determined by looking up the maximum principal permeability value of the permeability tensor of the weak surface in that area. It can be understood that the task scheduling optimization and sorting adopts a priority-based scheduler. The roof collaborative control trajectory derived from the collapse risk area has the highest priority, the dynamic grouting intervention strategy derived from the water hazard threat area has the second highest priority, and the mining trajectory replanning scheme derived from the resource inefficient area has the basic priority. See Table 1.

[0038] Table 1: Mapping Table for Adjusting Mining Process Parameters

[0039] Optionally, the mining sequence reorganization scheme is achieved by adjusting the reciprocating cutting sequence of the coal mining machine near resource-inefficient areas and water-threatened areas, inserting detour or priority mining instructions into the advancement plan generated by the intelligent decision kernel. It can be understood that the set of collaborative intervention instructions is ultimately encoded into standardized control instruction messages, which are then distributed through the industrial ring network to the electro-hydraulic control system of the working face hydraulic support, the coal mining machine control system, and the grouting station pumping control system.

[0040] See Figure 4 This is a dual-axis dynamic monitoring chart of roof control during the mining process of a gold mine, showing the changing trends of two key indicators within mining cycles 1-10. The hydraulic support strength gradually increases from 20 MPa in cycle 1, reaching a peak of 45 MPa in cycle 5, and then gradually decreases to 28 MPa in cycle 10. This reflects the dynamic change of roof pressure during the face advancement process, with cycle 5 being a critical stage of roof stress concentration. The difference in support advance step distance is highly synchronized with the trend of support strength, reaching a maximum of 1.0 m in cycle 5, and then decreasing as the support strength decreases. The change in step distance difference indicates that the system optimizes the roof stress distribution and reduces the risk of collapse by adjusting the advance step distance of adjacent supports during the stress concentration stage.

[0041] In one embodiment of the present invention, in a specific implementation, the adaptive mining system for gold mine working faces based on digital twins performs model verification and parameter calibration. The implementation method includes, within a set model verification time window, simultaneously collecting measured roof pressure distance data and actual ore recovery rate data of the mining working face; performing a difference quantification analysis between the measured roof pressure distance data and the pressure distance distribution predicted by the digital twin geological body to generate a pressure distance prediction deviation correction factor; evaluating the regional consistency between the actual ore recovery rate data and the predicted recovery rate derived from the composite geological characterization to generate a recovery rate prediction deviation correction factor; using the pressure distance prediction deviation correction factor and the recovery rate prediction deviation correction factor to perform parameter adaptive calibration of the evaluation logic of the intelligent decision kernel to generate a calibrated intelligent decision kernel; and deploying the calibrated intelligent decision kernel to apply to the adaptive decision-making process of mining in future working faces.

[0042] In some embodiments, the model validation time window is set to the cycle of completing one full coal mining machine cutting cycle. The measured roof pressure step distance data is obtained through a pressure sensor array arranged on the hydraulic support. The pressure sensor array records periodic pressure peaks and their corresponding spatial positions. The measured roof pressure step distance data is calculated from the distance between adjacent pressure peak locations. In specific implementations, the actual ore recovery rate data is calculated comprehensively using weighing sensors and belt scales installed on the scraper conveyor, combined with the grade analysis results of the ore. It can be understood that the pressure step distance distribution predicted by the digital twin geological body is based on the quantification field of rock mass damage and the envelope surface of critical instability stress, and is a map of periodic roof fracture locations obtained through mechanical calculation simulation. The predicted recovery rate derived from the composite geological characterization is a theoretical recovery rate value obtained by combining the continuous topological structure of the ore body with preset mining process parameters, through simulation calculation using a three-dimensional block model and mining process.

[0043] Optionally, the difference quantification analysis uses the root mean square error method to calculate the overall deviation between the predicted and measured values ​​on the spatial grid, in order to reduce the step prediction deviation correction factor. It is obtained by processing with a normalization function, and its mathematical expression is:

[0044] in: This represents the step distance prediction deviation correction factor, which is a dimensionless correction coefficient. This represents the total number of pressure events collected within the model validation time window. Representing the digital twin geological body to the first The pressure step size predicted by the next pressure event. This represents the first value obtained by actual measurement and calculation using a pressure sensor array. The pressure step distance value of the next pressure event. It can be understood that regional conformity assessment divides the mining area into multiple assessment units, compares the actual ore recovery rate with the predicted recovery rate within each assessment unit, and uses a recovery rate prediction deviation correction factor. It is the weighted average of the conformity of all assessment units, with the weight determined by the proportion of ore reserves within each assessment unit.

[0045] In some embodiments, the evaluation logic of the intelligent decision kernel undergoes adaptive parameter calibration, specifically by adjusting the future step distance prediction deviation correction factor. As a weight, it is multiplied into the stability coefficient threshold determination formula used in the damage assessment stage of the intelligent decision-making kernel, and the recovery rate prediction deviation correction factor is applied. As a weight, it is multiplied into the economic parameters of the ore lean-loss ratio prediction model used in the resource optimization mining stage. In specific implementation, after generating the calibrated intelligent decision kernel, the system will automatically back up the original kernel parameters and load the new calibrated intelligent decision kernel parameter set. Optionally, when deploying the calibrated intelligent decision kernel to a future working face, the system will perform a feedforward calculation on the initial geological data of the working face to verify the rationality of the output of the calibrated kernel under the new data conditions before putting it into the formal adaptive decision-making process. It can be understood that the model verification and parameter calibration process is triggered cyclically according to a preset period, or manually triggered when there are significant changes in the mining geological conditions, to ensure that the prediction and decision-making capabilities of the intelligent decision kernel continuously adapt to the actual mining environment.

[0046] See Figure 5 This is a spatial distribution cloud map of gold ore body grades, showing the variation in gold grade (g / t) within the strike distance (-40m to 40m) and dip distance (-30m to 30m). The central area of ​​the map (strike distance 0m to 20m, dip distance -10m to 10m) shows a distinct red-orange cluster, with gold grades reaching up to 7g / t, representing the main enrichment area of ​​the ore body. At strike distances of -20m to 0m, dip distances of 20m to 30m, and -20m to -30m, there are orange-yellow areas with grades of approximately 5-6g / t, representing secondary enrichment zones. The blue areas in the map (such as the left and right edges) have grades of only 2-3g / t, belonging to low-grade ore or surrounding rock areas. This cloud map is key input data for digital twin geological bodies, directly supporting subsequent mining decisions. It can be used to calculate the effective thickness variation function and strike variation rate of the ore body, providing a basis for replanning the mining trajectory.

[0047] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A digital twin-based adaptive mining system for gold mine working faces, characterized in that, The system includes: The geological dynamic perception module acquires the full-cycle geological dynamic dataset generated by the mining face during mining activities. The full-cycle geological dynamic dataset includes the evolution trajectory of the ore body's original stress field, the spatial variation curve of grade, and the response data of the surrounding rock structure. The geological modeling module drives the construction of a hierarchically coupled digital twin geological body through the full-cycle geological dynamic dataset. The digital twin geological body is expressed as a composite geological characterization that includes a rock mass damage quantification field, a continuous topological structure of the ore body, and a permeability tensor of tectonic weak surfaces. The mining optimization module performs closed-loop optimization calculations on mining behavior and geological conditions based on the composite geological characterization, and extracts dynamic configuration schemes for mining process parameters and geological disturbance risk zoning maps during the working face advancement process. The time-based correction module performs long-term mechanical feedback correction on the dynamic configuration scheme of the mining process parameters based on the time-based damage mechanism of the rock mass, and outputs the time-based corrected dynamic configuration scheme of the mining process parameters. The long-term mechanical feedback correction is related to the matching relationship between the evolution trajectory of the original stress field of the ore body and the rheological properties of the rock mass. The collaborative intervention module uses the geological disturbance risk zoning map to derive a set of collaborative intervention instructions for the mining process.

2. The adaptive mining system for gold mine working faces based on digital twins according to claim 1, characterized in that, The construction of a hierarchically coupled digital twin geological body driven by the full-cycle geological dynamic dataset includes: The evolution trajectory of the original stress field of the ore body is deconstructed into multiple stress state segments according to the mining time sequence, and each stress state segment corresponds to a specific mining cycle. For each stress state segment, the following steps are performed: The surrounding rock structure response data is integrated to generate a multi-dimensional deformation collaborative network of the working face surrounding rock system. The multi-dimensional deformation collaborative network contains a roof delamination distribution cloud map, a two-sided convergence rate field and a floor activation intensity spectrum. The multidimensional deformation collaborative network is subjected to deep spatiotemporal correlation analysis with the current stress state segment to form a geological state snapshot corresponding to the current mining cycle. The geological state snapshot encapsulates the rock mass damage evolution tensor, the geometric description of the ore body mining boundary, and the water-conducting fracture development trend matrix. A progressive overlay evolutionary deduction was performed on multiple geological state snapshots arranged in chronological order to calculate the quantitative field of rock mass damage, the continuous topological structure of the ore body, and the permeability tensor of tectonic weak surfaces; among which... The rock mass damage quantification field is defined as the result of the convolution operation of the cumulative damage gradient along the mining direction of the rock mass damage evolution tensor and the energy dissipation rate of the deformation cooperative network. The continuous topological structure of the ore body is expressed as a combination of connectivity measures and morphological persistence features in three-dimensional space for the geometric description of the ore body's mining boundary. The permeability tensor of the structural weak surface is characterized as the anisotropic distribution of permeability within a unit geological volume element, representing the development trend matrix of water-conducting fractures.

3. The adaptive mining system for gold mine working faces based on digital twins according to claim 2, characterized in that, The step of performing deep spatiotemporal correlation analysis between the multi-dimensional deformation collaborative network and the current stress state segment to form a geological state snapshot corresponding to the current mining cycle includes: Based on the dependence between the delamination distribution cloud map of the top plate and the principal stress direction in the current stress state segment, a delamination principal stress coupled response model is established. The spatial distribution function of the rock mass damage evolution tensor is obtained by iteratively solving the delamination principal stress coupled response model. Based on the correlation mapping between the convergence rate field of the two sides and the tectonic stress concentration coefficient, a dynamic evaluation framework for the stability of the two sides is constructed. The dynamic evaluation framework for the stability of the two sides incorporates the real-time attenuation parameters of the rock mass strength softening coefficient and the shear strength of the structural surface. Combining the activation intensity spectrum of the base plate with the triggering conditions of water-rock interaction, a multi-step judgment logic for the risk of water inrush on the base plate is formed. The multi-step judgment logic incorporates a bidirectional feedback mechanism between pore water pressure propagation and fracture propagation rate. The spatial distribution function, the output of the dynamic evaluation framework for stability of the two sides, and the conclusion of the multi-step judgment logic for the risk of water inrush in the bottom plate are assimilated and fused into three-dimensional field data to generate a high-fidelity geological field data set that integrates damage state, ore body geometry and hydraulic characteristics.

4. The adaptive mining system for gold mine working faces based on digital twins according to claim 1, characterized in that, The closed-loop optimization calculation based on the composite geological characterization of mining behavior and geological conditions, extracting dynamic configuration schemes for mining process parameters and geological disturbance risk zoning maps during the working face advancement process, includes: The quantitative field of rock mass damage is imported into the damage assessment stage of the pre-set intelligent decision kernel, and the spatial coordinate set of potential collapse zone and critical instability stress envelope surface are identified by the instability state projection module. The continuous topology of the ore body is imported into the resource optimization mining stage of the intelligent decision kernel, and a mining process adaptability assessment is performed to generate a mining continuity index and a predicted spectrum of ore lean ratio. The permeability tensor of the constructed weak surface is imported into the water hazard early warning link of the intelligent decision kernel, and the evolution direction of the dominant water-conducting channel and the inrush risk intensity value are calculated based on the seepage mutation identification model. By integrating the critical instability stress envelope, the ore lean loss ratio prediction spectrum, and the water inrush risk intensity value, a comprehensive disturbance index for the working face is constructed. Based on the comparison between the comprehensive disturbance index of the working face and the dynamic safety boundary, a dynamic configuration scheme for mining process parameters is determined. Based on the geometric spatial clustering results of the spatial coordinate set, the mining continuity index, and the evolution direction of the dominant water-diverting channel, the spatial boundaries of the landslide risk zone, the resource inefficiency zone, and the water hazard threat zone are drawn.

5. The adaptive mining system for gold mine working faces based on digital twins according to claim 1, characterized in that, The method involves performing long-term mechanical feedback correction on the dynamic configuration scheme of mining process parameters based on the time-dependent damage mechanism of the rock mass, and outputting a time-dependent corrected dynamic configuration scheme of mining process parameters, including: The dominant stress evolution path and stress relaxation time constant in the evolution trajectory of the original stress field of the ore body are analyzed, and the time-varying reduction of the long-term strength parameters of the rock mass due to mining disturbance is calculated. Based on the time-varying reduction, stress history compensation calculation is performed on the rock mass damage quantification field to generate a time-varying corrected rock mass damage quantification field. By associating the stress relaxation time constant with the constitutive relationship of rock mass creep damage, creep fracture correction deduction is performed on the continuous topology of the ore body to generate a time-varying corrected ore body continuous topology. Based on the cumulative data of rock mass fatigue damage under the dominant stress path, the permeability tensor of the structural weak surface is subjected to fatigue damage coupling adjustment processing to generate a time-varying corrected permeability tensor of the structural weak surface. The time-varying modified rock mass damage quantification field, ore body continuity topology, and tectonic weak surface permeability tensor are re-input into the intelligent decision kernel for iterative calculation, and the dynamic configuration scheme of mining process parameters after long-term mechanical feedback correction is output.

6. The adaptive mining system for gold mine working faces based on digital twins according to claim 5, characterized in that, The step of performing stress history compensation calculation on the rock mass damage quantification field based on the time-varying reduction to generate a time-varying corrected rock mass damage quantification field includes: The design time-history strength degradation function is constructed by calling the rock mass strength benchmark parameter set of the working face under the original geostress environment and the time-varying reduction. The cumulative damage factor of the rock mass is calculated using the design time-history strength degradation function. The cumulative damage factor of the rock mass is characterized as the product of the stress path integral and the time-varying rate of the strength parameter. The cumulative damage factor of the rock mass is injected into the generation process of the rock mass damage quantification field to form a correction amount of the rock mass damage quantification field containing the historical damage accumulation effect. The rock mass damage quantification field correction is applied to compensate for the rock mass damage self-healing effect, which depends on the synergistic effect of fracture closure rate and time healing factor.

7. The adaptive mining system for gold mine working faces based on digital twins according to claim 1, characterized in that, The set of collaborative intervention instructions for the mining process derived from the geological disturbance risk zoning map includes: Facing the spatial boundary of the collapse risk zone, the optimal roof collaborative control trajectory is calculated. The optimal roof collaborative control trajectory is achieved by redistributing the support strength and step distance of the hydraulic support group. To address the spatial boundaries of the resource-inefficient zone, a replanning scheme for the mining trajectory is designed. This scheme includes curvature optimization of the adaptive cutting path of the coal mining machine and load matching adjustment of the cutting power. Based on the spatial boundaries of the water hazard threat zone, a dynamic grouting intervention strategy is formulated. The dynamic grouting intervention strategy dynamically matches the grouting timing and grout diffusion radius based on the changing trend of the water inrush risk intensity value. The optimal roof collaborative control trajectory, the mining trajectory replanning scheme, and the dynamic grouting intervention strategy are optimized and sorted by task scheduling to form a set of collaborative intervention instructions containing execution sequences and control thresholds; The set of collaborative intervention instructions for the mining process is specified as a strategy for adjusting the intensity of roof maintenance and a scheme for reorganizing the mining sequence.

8. The adaptive mining system for gold mine working faces based on digital twins according to claim 7, characterized in that, The design forms a replanning scheme for the mining trajectory, including: Extract the spatial morphological features of the resource inefficiency zone and calculate the effective thickness variation function and strike variation rate of the ore body; The rotation speed control strategy of the coal mining machine drum is dynamically adapted based on the effective thickness variation function of the ore body, and the rotation speed control strategy is synchronized with the local extreme value change trend of the effective thickness. The stability control law of the coal mining machine traction mechanism is optimized in real time based on the aforementioned change rate of trajectory, so that the fluctuation amplitude of the traction acceleration is constrained to be proportional to the reciprocal of the change rate of trajectory. The cutting depth of the cutting teeth is dynamically configured based on the hardness identification results of the rock-ore interface to ensure that the cutting ratio energy consumption is within the preset economic operating range. A mapping table for adjusting mining process parameters is generated, which integrates rotation speed control strategy, stability control law parameters, and cutting depth configuration.

9. The adaptive mining system for gold mine working faces based on digital twins according to claim 4, characterized in that, The calculation of the evolution direction of the dominant water-conducting channel and the inrush risk intensity value based on the seepage mutation identification model includes: Obtain the data sequence of the permeability tensor of the constructed weak surface under multiple consecutive mining time series; Calculate the permeability variation gradient field of each geological element between adjacent time series based on the data sequence; Identify connected regions in the permeability change gradient field where the gradient value exceeds a preset threshold, and mark each connected region as a potential water-conducting channel; Calculate the spatial displacement vector of the geometric center point of each potential water-conducting channel between the current time series and the previous time series, and determine the direction of the spatial displacement vector as the current evolution direction of the potential water-conducting channel; The maximum gradient value of the permeability change gradient field in each potential water-conducting channel is extracted, and combined with the volume of the potential water-conducting channel, the initial value of the inrush risk intensity of each potential water-conducting channel is obtained through weighted fusion calculation. Based on the rock mechanics properties reflected by the permeability tensor of the structural weak surface, the initial value of the inrush risk intensity of each potential water-conducting channel is corrected for rock stability, and the inrush risk intensity value of each potential water-conducting channel is output.

10. The adaptive mining system for gold mine working faces based on digital twins according to claim 1, characterized in that, The system also includes: Within the set model verification time window, the measured roof pressure step distance data and actual ore recovery rate data of the mining face are collected synchronously. The measured data of the pressure step distance of the roof is compared with the pressure step distance distribution predicted by the digital twin geological body to generate a pressure step distance prediction deviation correction factor. The actual ore recovery rate data and the predicted recovery rate derived from the composite geological characterization are evaluated for regional consistency, and a recovery rate prediction deviation correction factor is generated. The evaluation logic of the intelligent decision kernel is adaptively calibrated using the aforementioned step distance prediction deviation correction factor and recovery rate prediction deviation correction factor to generate a calibrated intelligent decision kernel. The calibrated intelligent decision kernel will be deployed and applied to the adaptive decision-making process for mining in future working faces.

Images