An intelligent early warning system for stope roof fracture based on hard rock damage evolution

Abstract

The present invention relates to the field of rock mechanics monitoring technology, and in particular discloses an intelligent early warning system for stope roof breakage based on hard rock damage evolution. The system comprises: a geotechnical property detection module, which acquires geotechnical property parameters of the rock and soil beneath the stope roof based on measurement while drilling during drilling, including stress response, wave velocity change, crack evolution, and porosity; a drilling data acquisition module, which acquires drilling process data including bit weight, torque, and drilling speed during drilling, and which includes a sensor array connected to the drilling equipment; and a data processing module, which, based on the geotechnical property parameters and drilling process data and a hard rock damage evolution algorithm, obtains analysis results associated with the damage state of the stope roof, and determines the breakage risk of the stope roof based on the analysis results. By integrating geotechnical properties with drilling data in real time and applying a hard rock damage evolution algorithm, the system provides early warning of stope roof breakage risk.

Description

Technical Field

[0001] The present invention relates to the technical field of mining engineering and rock mechanics monitoring, and in particular to an intelligent early warning system for stope roof breakage based on hard rock damage evolution. Background Art

[0002] With the continuous advancement of mining technology, the depth and scale of mines have gradually increased. Stope stability has become a key factor in mine safety. The stability of the stope roof directly affects the safe operation of the mine. Roof failure is particularly common in hard rock conditions. Once a stope roof fails in a hard rock geological environment, it not only interrupts mine production but also poses a threat to the lives of miners.

[0003] Currently, traditional methods for monitoring stope roof stability rely primarily on manual inspections and limited sensor data collection. While these methods can monitor roof deformation and damage to a certain extent, their real-time monitoring and accuracy remain significant shortcomings. Traditional technologies primarily rely on simple monitoring of roof deformation and geological characteristics to infer potential failure risks, often failing to accurately and timely assess the damage status of the stope roof.

[0004] Traditional monitoring systems have a certain lag in data collection, making them unable to reflect dynamic changes in the mine roof in real time. Furthermore, existing technologies are often limited to a single data source, resulting in weak early warning capabilities for roof failures, making it difficult to accurately identify potential risks at an early stage. Summary of the Invention

[0005] In view of this, an embodiment of the present invention provides an intelligent early warning system for stope roof fracture based on hard rock damage evolution to solve at least one of the above technical problems.

[0006] To achieve the above objectives, in a first aspect, an intelligent early warning system for stope roof fracture based on hard rock damage evolution is provided, which comprises:

[0007] A geotechnical property detection module is used to obtain geotechnical property parameters of the rock and soil below the stope roof based on measurement while drilling during drilling. The geotechnical property parameters include stress response parameters, wave velocity change parameters, crack evolution parameters, and porosity parameters of the rock and soil;

[0008] A drilling data acquisition module, configured to acquire drilling process data including weight on bit, torque, and drilling speed during the drilling process. The drilling data acquisition module includes a sensor array connected to the drilling equipment for real-time monitoring and recording of the drilling process data.

[0009] A data processing module is used to obtain analysis results associated with the damage state of the stope roof based on the geotechnical physical property parameters and the drilling process data in combination with a hard rock damage evolution algorithm, and to determine the fracture risk of the stope roof based on the analysis results.

[0010] The above technical solution has the following beneficial technical effects:

[0011] The aforementioned intelligent early warning system for mine roof fracture based on hard rock damage evolution integrates geophysical property detection, drilling data acquisition, and data processing modules, combined with a hard rock damage evolution algorithm, to achieve comprehensive intelligent monitoring and early warning of the mine roof. First, the geophysical property detection module uses measurement while drilling technology to acquire multiple key physical parameters, including geophysical stress response, wave velocity changes, crack evolution, and porosity, comprehensively reflecting the physical properties and changes of the geotechnical layer. Second, the drilling data acquisition module monitors drilling pressure, torque, drilling speed, and other data in real time during the drilling process, accurately recording the operating status of the drilling equipment and changes in the geological environment, ensuring comprehensive and efficient data acquisition. Finally, the data processing module integrates geophysical property parameters with drilling process data, analyzes them using the hard rock damage evolution algorithm, and, combining complex geological conditions with the mechanical properties of the drilling process, provides real-time assessment of the mine roof damage status. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] The accompanying drawings are provided for a better understanding of the present invention and are not intended to limit the present invention.

[0013] Figure 1 This is a structural block diagram of an intelligent early warning system for stope roof fracture based on hard rock damage evolution in an embodiment of the present invention;

[0014] Figure 2 This is a structural block diagram of another intelligent early warning system for stope roof fracture based on hard rock damage evolution in an embodiment of the present invention;

[0015] Figure 3 This is a schematic diagram of the first working principle of the geotechnical physical property detection module in an embodiment of the present invention;

[0016] Figure 4 Schematic diagram of the second working principle of the geotechnical physical property detection module in an embodiment of the present invention;

[0017] Figure 5 Schematic diagram of the third working principle of the geotechnical physical property detection module in an embodiment of the present invention;

[0018] Figure 6 2 is a schematic diagram of the fourth working principle of the geotechnical physical property detection module in an embodiment of the present invention;

[0019] Figure 7This is a flow chart of an intelligent early warning method for stope roof fracture based on hard rock damage evolution in an embodiment of the present invention;

[0020] Figure 8 It is a schematic diagram of the structure of a computer system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0021] The following description of exemplary embodiments of the present invention is made in conjunction with the accompanying drawings, in which various details of the embodiments of the present invention are included to facilitate understanding. These details should be considered as merely exemplary. Therefore, it should be appreciated by those skilled in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0022] like Figure 1 As shown, an intelligent early warning system for stope roof fracture based on hard rock damage evolution includes:

[0023] A geotechnical property detection module is used to obtain geotechnical property parameters of the rock and soil below the stope roof based on measurement while drilling during drilling. The geotechnical property parameters include stress response parameters, wave velocity change parameters, crack evolution parameters, and porosity parameters of the rock and soil;

[0024] A drilling data acquisition module, configured to acquire drilling process data including weight on bit, torque, and drilling speed during the drilling process. The drilling data acquisition module includes a sensor array connected to the drilling equipment for real-time monitoring and recording of the drilling process data.

[0025] A data processing module is used to obtain analysis results associated with the damage state of the stope roof based on the geotechnical physical property parameters and the drilling process data in combination with a hard rock damage evolution algorithm, and to determine the fracture risk of the stope roof based on the analysis results.

[0026] In this embodiment, the geotechnical property detection module primarily relies on high-precision sensing devices integrated into the drilling equipment to implement measurement while drilling. Specifically, stress sensors, sonic velocimeters, fracture imaging probes, and porosity detectors are installed on the drill bit and drill rod surfaces of the drilling equipment at predetermined intervals and layouts. The stress sensors are piezoresistive sensors that can sense the stress effects of the rock and soil on the drill bit and drill rod in real time during drilling, thereby obtaining rock and soil stress response parameters such as the magnitude and direction of principal stresses. The sonic velocimeter measures the propagation velocity of elastic waves in the rock and soil by transmitting and receiving elastic waves, thereby obtaining a wave velocity variation parameter that reflects the density and damage state of the rock and soil. The fracture imaging probe is equipped with a high-resolution camera and light source, which can image and record rock and soil fractures exposed during drilling, thereby analyzing fracture evolution parameters such as the degree of development, length, width, and distribution of the fractures. The porosity detector uses the principles of gas permeation or pressure decay to measure the ratio of pore volume to total volume in the rock and soil to obtain porosity parameters. These sensors work synchronously with the drilling equipment. During the drilling process, as the drill bit advances, they collect various physical properties of the rock and soil below the mine roof in real time, and transmit the collected data in real time to the subsequent data processing module via wired transmission (such as optical fiber) or wireless transmission (such as Bluetooth, WiFi).

[0027] The drilling data acquisition module consists of an array of sensors connected to key components of the drilling equipment, used to monitor and record key data during the drilling process in real time. Specifically, a pressure sensor is installed on the drilling equipment's weight-on-bit application mechanism (such as a hydraulic or pneumatic cylinder). This pressure sensor accurately measures the weight-on-bit applied to the drill bit. A torque sensor is installed on the output shaft of the rotary drive mechanism (such as a motor and gearbox) to detect the torque applied to the drill bit during drilling. Drilling speed is measured using a displacement sensor (such as a linear displacement sensor or encoder) installed on the drilling equipment's feed mechanism. This sensor records the drill rod's feed displacement per unit time in real time, thereby calculating the drilling speed. These sensors are all industrial-grade, high-precision sensors with strong anti-interference capabilities and high reliability. The sensor array is connected to the control system of the drilling equipment through a data acquisition card. The data acquisition card collects drilling process data such as drilling pressure, torque, and drilling speed in real time at a set sampling frequency (for example, 100 times per second), converts the collected data into digital signals, and transmits them to the data processing module via industrial Ethernet or other high-speed data transmission buses for subsequent analysis and processing.

[0028] The data processing module's primary function is to comprehensively process and analyze data transmitted from the geophysical property detection module and the drilling data acquisition module. Combined with the hard rock damage evolution algorithm, it generates analysis results related to the damage state of the stope roof and determines the risk of roof failure. First, the data processing module preprocesses the acquired geophysical property parameters and drilling process data, including data filtering, noise reduction, and normalization, to remove noise and outliers and ensure data accuracy and reliability. The preprocessed data is then input into the hard rock damage evolution algorithm. Based on the fundamental principles of rock mechanics and damage mechanics, the algorithm establishes a mathematical model linking geophysical property parameters, drilling process data, and hard rock damage evolution. By analyzing stress response parameters, wave velocity variation parameters, fracture evolution parameters, porosity parameters, as well as data such as bit weight, torque, and drilling speed, it calculates hard rock damage variables, such as damage degree and damage rate, thereby quantitatively describing the damage state of the stope roof. Next, the fracture risk of the stope roof is assessed based on the preset correspondence between damage status and fracture risk. For example, when the damage level exceeds a certain threshold, the stope roof is determined to be at high risk of fracture, and a corresponding warning signal is issued. Furthermore, the data processing module can store and analyze historical data, establishing a database for long-term monitoring and trend prediction of the damage evolution of the stope roof, providing a more reliable basis for safe mining in the stope. The data processing module can be implemented using an industrial control computer or dedicated data analysis equipment. Its software utilizes a modular design, offering excellent scalability and compatibility, enabling easy data exchange and integration with other monitoring systems.

[0029] In some specific embodiments, hard rock damage evolution algorithms may employ machine learning-based hard rock damage evolution algorithms, fractal theory-based hard rock damage evolution algorithms, or nonlinear hard rock damage evolution algorithms based on damage constitutive relations. Machine learning-based hard rock damage evolution algorithms use deep neural networks to predict and model the evolution of hard rock damage. Deep learning algorithms can learn complex nonlinear relationships from large amounts of drilling data, geophysical parameters, and environmental factors. In particular, convolutional neural networks and recurrent neural networks can effectively process both time series and spatial data, providing new predictive perspectives for hard rock damage evolution. Fractal-based damage evolution models employ fractal theory to study hard rock damage evolution. Rock cracks often exhibit self-similarities and multi-scale characteristics, and fractal geometry can be used to accurately describe the crack propagation process. Fractal dimension analysis can quantitatively describe the damage evolution of rocks under different loads. Nonlinear damage evolution algorithms based on damage constitutive relations can more accurately simulate rock damage processes based on nonlinear constitutive relations. By introducing variables such as damage degree and damage rate, a damage evolution model that is consistent with the actual behavior of rock materials can be established.

[0030] This embodiment uses a geotechnical physical property detection module to acquire multi-dimensional geotechnical characteristic parameters such as stress response, wave velocity changes, crack evolution, and porosity in real time. Combined with dynamic drilling process data such as drilling pressure, torque, and drilling speed collected simultaneously by the drilling data acquisition module, this system constructs a multi-source data acquisition system covering both hard rock physical properties and drilling behavior. The data processing module couples and analyzes these two types of data based on a hard rock damage evolution algorithm, accurately capturing the damage evolution characteristics of the stope roof under drilling disturbances and enabling cross-scale correlation analysis from microscopic crack development to macroscopic fracture risk. This solution avoids the limitations of traditional single-parameter monitoring and, through multi-source data fusion and intelligent algorithm drive, improves the real-time, comprehensiveness, and accuracy of stope roof fracture warnings.

[0031] like Figure 2 and Figure 3 As shown, in a further embodiment, the rock and soil physical property detection module includes a stress response measurement unit and a first processing unit; the stress response measurement unit is configured with a stress sensor array arranged along the drilling direction, and the stress sensor array is connected to the drill pipe of the drilling equipment, and is used to collect the original signals corresponding to the circumferential stress, axial stress and radial stress at different drilling depths; the first processing unit is used to calculate the stress response parameters of the rock and soil through a stress tensor solution model based on the original signals corresponding to the circumferential stress, axial stress and radial stress at different drilling depths.

[0032] The input to the stress tensor solution model is the raw signals corresponding to the hoop stress, axial stress, and radial stress at different drilling depths, collected by the stress sensor array. The stress tensor solution model outputs stress response results that are arranged sequentially along the depth of the borehole and have multidimensional characteristics. These results include principal stress amplitudes (including maximum principal stress, intermediate principal stress, and minimum principal stress), which clearly reflect the magnitude of stress within the rock and soil at different depths; maximum shear stress, which can be used to assess the risk of rock and soil failure under shear; average stress, which helps determine the overall stress state of the rock and soil; principal stress direction angles (presented as azimuth and inclination), which clarify the spatial direction of the principal stresses, thereby determining stress concentration areas and the direction of potential failure surfaces; and stress state types (such as tension, compression, or strike-slip), which can assist in determining the stress mode of the rock and soil.

[0033] The stress tensor solution model uses a multi-stage processing flow to calculate the stress response parameters of the geotechnical mass. Specifically, in the signal preprocessing step, the raw signal is first input into the signal preprocessing module for digital filtering-based signal denoising, temperature drift compensation, and dynamic baseline calibration. In the coordinate system conversion step, the preprocessed stress components are converted from the sensor's local coordinate system to the geodetic coordinate system to reflect the true stress state of the geotechnical mass. The first processing unit, combining the pitch, azimuth, and roll angle data collected by the drill pipe attitude sensor, projects the hoop, axial, and radial stress components in the local coordinate system into a global coordinate system using a spatial rotation matrix. This coordinate system is referenced to the vertical direction of the borehole, enabling spatial alignment of stress data at different depths. In the three-dimensional stress tensor construction step, a three-dimensional stress tensor representing the stress state of the geotechnical mass is constructed based on the orthogonal stress components in the global coordinate system. The off-diagonal components of the stress tensor are supplemented by integrating shear stress correlation data from adjacent sensor groups at the same depth. In the eigenvalue decomposition step, the first processing unit uses eigenvalue decomposition to extract the magnitudes and spatial directions of the three principal stresses from the three-dimensional stress tensor. The principal stress directions, described as azimuth and inclination angles, reflect the trends of the maximum, intermediate, and minimum principal stresses within the geotechnical mass, thereby determining the stress concentration areas and failure plane directions within the geotechnical mass.

[0034] In this embodiment, the stress response measurement unit comprises an array of stress sensors spaced evenly along the drilling direction and a supporting mechanical structure. Specifically, a triaxial stress sensor array is circumferentially mounted on the outer surface of the drilling equipment's drill pipe at intervals of 0.5 meters. Each array contains three orthogonally arranged resistance strain gauge sensors, each responsible for collecting stress components in the circumferential (tangential), axial (drilling direction), and radial (radial toward the borehole center) directions. The sensors are secured to pre-defined grooves in the drill pipe with a high-strength adhesive. The grooves are coated with a transparent, wear-resistant protective coating, ensuring that the sensors and the drill pipe are simultaneously subjected to geotechnical stresses while preventing damage to the sensors from friction caused by rock debris during drilling. The signal output of the stress sensor is connected to a miniature data acquisition module inside the drill pipe via a high-temperature shielded cable. This data acquisition module integrates signal amplification and analog-to-digital conversion functions, acquires raw stress signals in real time at a sampling frequency of 200 Hz, and transmits the data to a ground control system or data processing module via a pre-defined fiber optic bus or wireless transmission module (e.g., a low-speed wireless personal area network based on the IEEE 802.15.4 protocol) within the drill pipe. The layout design of the stress sensor array meets the requirements of three-dimensional stress state measurement. Its axial sensors are arranged along the axis of the drill pipe, the circumferential sensors are arranged along the tangent direction of the drill pipe circumference, and the radial sensors are perpendicular to the drill pipe surface and point to the center of the circle, ensuring that the original signals of the circumferential stress, axial stress and radial stress of the rock and soil on the drill pipe at different drilling depths can be accurately captured.

[0035] In this embodiment, the first processing unit receives the original signals of circumferential stress, axial stress and radial stress collected by the stress sensor array, and solves the stress response parameters of the rock and soil through a multi-stage processing flow. First, the original signal is input into the signal preprocessing module, and signal denoising, temperature drift compensation and dynamic baseline calibration based on digital filtering are performed. In order to deal with the high-frequency noise introduced during the transmission process, digital filtering technology is used to retain the effective frequency band signal and remove the high-frequency noise; at the same time, according to the real-time data of the built-in temperature probe of the sensor, the sensitivity deviation of the strain gauge caused by the change of ambient temperature is dynamically corrected to achieve temperature drift compensation. The dynamic baseline calibration is based on the zero stress reference value when the drill pipe is in a static state, eliminating the influence of the drill pipe's own weight and installation residual stress, and ensuring the accuracy of the measurement benchmark.

[0036] In this embodiment, the preprocessed stress components must be converted from the sensor's local coordinate system to the geodetic coordinate system to reflect the true stress state of the rock mass. The first processing unit, combining the pitch, azimuth, and roll angle data collected by the drill pipe attitude sensor, projects the hoop, axial, and radial stress components from the local coordinate system into a global coordinate system using a spatial rotation matrix. This coordinate system, based on the vertical direction of the borehole, spatially aligns stress data at different depths, providing a unified reference framework for subsequent three-dimensional stress analysis.

[0037] In this embodiment, a three-dimensional stress tensor representing the stress state of the geotechnical mass is constructed based on orthogonal stress components in a global coordinate system. The off-diagonal components of the stress tensor are supplemented by integrating shear stress correlation data from adjacent sensor groups at the same depth. The first processing unit uses eigenvalue decomposition to extract the magnitudes and spatial directions of the three principal stresses from the three-dimensional stress tensor. The principal stress directions, described as azimuths and inclinations, reflect the trends of the maximum, intermediate, and minimum principal stresses within the geotechnical mass, thereby determining stress concentration areas and the direction of the failure plane.

[0038] In this embodiment, the first processing unit outputs stress response results containing the following parameters: principal stress amplitudes (maximum principal stress, intermediate principal stress, minimum principal stress), maximum shear stress, mean stress, principal stress orientation angle, and stress state type (e.g., tension, compression, or strike-slip). These parameters are arranged sequentially by borehole depth, forming a continuous stress distribution dataset along the borehole profile. The processed results, after being verified for logical consistency by a data verification module, are transmitted to the ground control system, providing a direct basis for geotechnical analysis and engineering safety assessment.

[0039] In some alternative embodiments, the aforementioned stress tensor solution model can be replaced with a deep learning-based stress tensor solution model. This model uses a deep learning algorithm to directly establish a nonlinear mapping relationship between the raw stress sensor signals and the rock and soil stress tensor parameters. This model takes the raw circumferential, axial, and radial stress signals as input, and can also incorporate auxiliary data such as drilling speed, torque, and drill pipe attitude angle. After preprocessing (noise reduction and normalization), these signals are input into a trained neural network (such as a deep neural network, graph neural network, or convolutional neural network) or an ensemble learning model (such as a random forest or gradient boosting tree). The model is trained using a large amount of laboratory rock sample test data and field drilling data. During the learning process, the model automatically captures the complex coupling between multidimensional stress signals and output parameters such as principal stress amplitudes (maximum, intermediate, and minimum principal stresses), maximum shear stress, and principal stress orientation angles.

[0040] This embodiment realizes the synchronous acquisition of the circumferential stress, axial stress and radial stress of rock and soil at different depths through the stress sensor array arranged along the drilling direction, and can comprehensively obtain the three-dimensional stress distribution characteristics during the drilling process; combined with the stress tensor solution model, the multi-directional stress original signal is integrated into key parameters such as the principal stress and shear stress of the rock and soil, which not only improves the accuracy and spatial continuity of the stress state analysis, but also provides a multi-dimensional data basis for the subsequent rock and soil stability assessment, and enhances the real-time and reliability of the mine roof rupture warning.

[0041] like Figure 4 As shown, in a further embodiment, the geotechnical physical property detection module further includes a wave velocity detection unit and a second processing unit; the wave velocity detection unit includes an acoustic wave transmitting probe installed in the drill bit and an acoustic wave receiving array arranged in the middle of the drill pipe, the acoustic wave transmitting probe is used to transmit pulsed acoustic waves, and the acoustic wave receiving array is used to collect reflection signals and transmission signals of different interfaces and generate initial detection data, the initial detection data including signal arrival time, signal amplitude and phase information; wherein the signal amplitude refers to the intensity change of the reflection signal and the transmission signal, and the phase information refers to the phase difference between the reflection signal and the transmission signal;

[0042] The second processing unit is used to calculate the correspondence between the propagation time and propagation distance of the acoustic wave signal at different depths based on the initial detection data and the corresponding drilling depth coordinates and based on the time-distance curve fitting algorithm, and calculate the longitudinal wave velocity and transverse wave velocity of the geotechnical medium based on the correspondence and the physical properties of the geotechnical medium; the second processing unit is also used to obtain the wave velocity change parameter by performing difference calculation and gradient analysis on the longitudinal wave velocity and transverse wave velocity at adjacent depths.

[0043] Specifically, the wave velocity detection unit consists of an acoustic wave transmitter probe installed inside the drill bit and an acoustic wave receiver array located in the middle of the drill pipe. The acoustic wave transmitter probe is made of piezoelectric ceramic and encapsulated in a high-temperature resistant alloy housing. Its transmitting end face is flush with the drill bit cutting teeth, ensuring that the acoustic wave pulses are directly transmitted into the rock and soil medium ahead. The probe transmits a broadband acoustic wave signal with a center frequency of 50 kHz at a pulse frequency of 1 kHz, triggering a transmission every 0.1 meter of drilling depth. The acoustic wave receiver array consists of eight receiving sensors arranged evenly spaced along the drill pipe axis, with a spacing of 0.5 meters between adjacent sensors. Each sensor has a built-in high-sensitivity piezoelectric element and is tightly attached to the inner wall of the drill pipe via a preloaded spring to ensure efficient acoustic wave coupling. After the acoustic wave is transmitted, the receiving array initiates synchronous sampling, continuously recording the reflected signal (from the rock and soil interface) and the transmitted signal (through the rock and soil) at a sampling rate of 1 MHz, while simultaneously extracting the signal's arrival time, peak amplitude, and phase offset.

[0044] Specifically, the raw waveform data recorded by the acoustic wave receiving array undergoes data preprocessing to generate initial detection data. The signal arrival time is determined using a threshold detection method, where the arrival time is marked when the signal amplitude first exceeds three times the root mean square value of the background noise. The signal amplitude is characterized by calculating the peak-to-peak values ​​of the reflected and transmitted waves, with variations in their intensity reflecting the differences in acoustic impedance at the interfaces between different rock and soil layers. Phase information is extracted using a Hilbert transform to determine the instantaneous phase angle, the difference of which is used to identify the interface reflection coefficient and the morphology of the rock interface (e.g., flat, broken, or tilted). Data is stored by drilling depth coordinate index, with each depth point associated with a corresponding transmit-receive distance, amplitude, phase, and time tag.

[0045] Specifically, the second processing unit uses a time-distance curve fitting algorithm based on the initial detection data to calculate the longitudinal wave velocity (Vp) and shear wave velocity (Vs) of the rock and soil. For each drilling depth, a time-distance curve is plotted, with the distance between the acoustic wave emission point and each receiving sensor as the horizontal axis and the corresponding signal arrival time as the vertical axis. The slope of the curve is fitted using the least squares method, yielding the longitudinal wave propagation velocity Vp = ΔL / Δt (ΔL is the incremental distance between the emission and reception, and Δt is the incremental time). The shear wave velocity Vs is calculated by separating the shear wave component from the waveform: using the polarization difference between the reflected and transmitted signals, combined with the azimuth angle data of the sensors in the receiving array, the shear wave arrival time is identified and a similar fit is performed. The physical properties of the rock and soil (e.g., density) are correlated using a pre-set empirical relationship table between the longitudinal wave velocity (Vp), shear wave velocity (Vs), and density to further verify the rationality of the velocity calculation results.

[0046] Specifically, the second processing unit performs difference calculations on the Vp and Vs values ​​at adjacent depth points (0.1 meter apart) in the same borehole to obtain velocity changes ΔVp and ΔVs. Simultaneously, a sliding window gradient analysis is performed, using a window length of 1 meter and a sliding average method to calculate the velocity gradient (i.e., the rate of change of velocity with depth). The formula is gradient = ΔV / ΔH (ΔV is the velocity range within the window, and ΔH is the window height). When ΔVp or ΔVs exceeds a preset threshold (e.g., ΔVp > 200 m / s) or the gradient value is abnormal (e.g., gradient > 500 m / s / m), it is determined that rock damage or structural surface development exists in that depth segment. The processing results are output as a velocity change parameter set, including velocity differences, gradient values, and anomaly markers, which are used to quantify the heterogeneity and damage degree of the rock and soil mass.

[0047] Specifically, by analyzing velocity variation parameters and combining them with a hard rock damage evolution model, potential damage areas in the stope roof can be identified. For example, a sudden drop in the longitudinal wave velocity (Vp) (e.g., from 4500 m / s to 3000 m / s) indicates the presence of cracks or cavities in the rock mass; an abnormal gradient in the shear wave velocity (Vs) reflects the distribution of bedding planes or dense joint zones.

[0048] The advantage of this technical solution is that by integrating the acoustic wave transmitting probe and the receiving array, it realizes the multi-dimensional collection of the acoustic wave reflection and transmission signals of the rock and soil during the drilling process, and can synchronously obtain the acoustic wave arrival time, amplitude and phase information, and combine the time-distance curve fitting algorithm to accurately solve the longitudinal and shear wave velocities, and quantitatively characterize the elastic properties of the rock and soil medium; further, through the difference and gradient analysis of the wave velocities at adjacent depths, it can effectively identify the wave velocity anomaly areas caused by rock damage or structural surface development, and provide dynamic wave velocity parameter support for the hard rock damage evolution analysis.

[0049] like Figure 5 As shown, in a further embodiment, the geotechnical physical property detection module also includes a crack evolution monitoring unit and a third processing unit; the crack evolution monitoring unit is used to obtain a hole wall image through a camera integrated in the side wall of the drill pipe; the third processing unit is used to use a digital image recognition algorithm to perform edge detection and contour tracking on the cracks in the hole wall image, quantify the spatial morphological parameters of the cracks and the connection relationship between the crack endpoints to obtain the connectivity of the cracks, calculate the crack density according to the number of cracks within the unit hole wall length, and calculate the crack fractal dimension according to the fractal box dimension algorithm of the crack geometric distribution and the connectivity; the crack evolution parameters include the crack density and the crack fractal dimension.

[0050] Specifically, the fracture evolution monitoring unit may include a high-definition camera assembly integrated into the sidewall of the drill pipe. The camera utilizes an impact-resistant, dustproof, and waterproof industrial-grade design and is embedded in an annular groove on the outer wall of the drill pipe. The groove surface is covered with a transparent tempered glass protective cover to protect it from rock debris during drilling. The camera lens axis is perpendicular to the borehole wall and is equipped with a ring-shaped fill light to provide uniform illumination in low-light downhole environments. The camera captures a panoramic image of the borehole wall at a frequency of 4096×2160 pixels, covering a 360-degree circumference of the borehole, triggered every 0.2 meters of drilling. A depth-synchronized trigger mechanism is employed, whereby the capture trigger signal is synchronously controlled by the drilling depth sensor, ensuring that the image is precisely correlated with the drilling depth coordinates. This depth-synchronized trigger mechanism ensures that the captured image data matches the depth information obtained during drilling. Image data is uploaded to the ground control system or data processing module in real time via a shielded cable within the drill pipe or a wireless transmission module.

[0051] Specifically, after the original borehole wall image is input into the third processing unit, it first undergoes image preprocessing. This includes contrast enhancement based on histogram equalization to highlight the grayscale difference between cracks and intact rock mass; a median filter to eliminate local noise caused by dust adhesion; and a cylindrical projection transformation to correct image distortion and expand the annular borehole wall image into a flat rectangular image. In the preprocessed image, cracks appear as dark linear features, while the background rock mass appears as a bright, uniform area, providing a clear input for subsequent identification.

[0052] Specifically, the third processing unit calls a digital image processing algorithm library to perform fracture identification on the preprocessed image. First, the Canny edge detection algorithm is used to extract fracture contours. A dual-threshold filter (a high threshold is used for strong edges, and a low threshold is used for weak edge connections) is used to preserve continuous fracture boundaries. Subsequently, a contour tracing algorithm is executed: starting from the upper left corner of the image, a pixel-by-pixel scan is performed. When an edge point is detected, boundary tracing is initiated, the coordinate sequence of the fracture contour is recorded, and the intersection points of adjacent fracture branches are marked. Contour data is stored in a vector polygon format, containing the length, average width, strike angle (relative to the borehole axis), and endpoint coordinates of each fracture.

[0053] Specifically, fracture density is calculated based on the number of fractures per unit borehole wall length. The third processing unit divides the borehole wall image into 0.2-meter segments based on the drilling depth coordinates. The number of fracture outlines within each segment is counted and divided by the segment length (0.2 meters) to obtain the fracture density value (unit: fractures / meter). Fracture connectivity is determined by analyzing the spatial relationship between adjacent fracture endpoints: if the horizontal projection distance between two fracture endpoints is less than a preset threshold (e.g., 5 mm) and the strike angle is less than 30 degrees, the fracture is considered potentially connected. The ratio of the number of connected fracture groups to the number of independent fractures is calculated as the connectivity indicator.

[0054] Specifically, the fractal dimension is used to quantify the complexity of the geometric distribution of fractures. The third processing unit uses the box counting method to calculate the fractal box dimension: the hole wall image is divided into multiple grids (the initial grid size is 1 / 10 of the image width, and it is gradually reduced to 1 / 100), and the number of grids N containing fracture pixels at each grid size is counted. A scatter plot is drawn with log(N) as the vertical coordinate and log(1 / grid size) as the horizontal coordinate. The slope of the linear regression fitting line is the fractal dimension. The fractal dimension is further weighted and corrected in combination with the fracture connectivity index. That is, if the connectivity is higher than 0.6, the weight coefficient is increased (for example, 1.2) to reflect the significant impact of the connected fracture network on the stability of the rock mass.

[0055] Specifically, the third processing unit outputs a set of fracture evolution parameters, including segmented fracture density, fractal dimension, and connectivity indicators, stored as a structured data table by depth sequence. This data, along with stress response parameters and wave velocity variation parameters, is synchronously transmitted to the early warning decision module. Through multi-parameter fusion analysis (for example, setting a fracture density greater than 10 fractures / meter and a fractal dimension greater than 1.5 as high-risk thresholds), the module dynamically assesses the cumulative degree of roof rock damage and triggers a graded early warning signal.

[0056] The advantage of this technical solution is that it directly obtains high-definition images of the hole wall through the camera integrated in the side wall of the drill pipe, and automatically identifies and quantifies the crack density, fractal dimension and connectivity parameters by combining digital image algorithms, thereby realizing non-contact, high-precision dynamic monitoring of the geometric characteristics of rock cracks; fractal dimension calculation combined with connectivity analysis can objectively characterize the spatial complexity and expansion trend of the crack network, which forms a multi-dimensional damage assessment system with stress parameters and wave velocity parameters, providing more comprehensive crack evolution data support for the early warning of mine roof rupture.

[0057] like Figure 6 As shown, in a further embodiment, the rock and soil physical property detection module also includes a porosity calculation unit, which is used to calculate the porosity parameter by combining the longitudinal wave velocity and transverse wave velocity measured by the wave velocity detection unit, and the rock and soil density obtained by the density sensor of the drilling rig.

[0058] In this embodiment, the porosity calculation unit includes a data interface for the drill rig's density sensor, a data interface for the velocity detection unit, and a porosity calculation model component. The density sensor, a gamma-ray densitometer, is integrated into the measuring sub at the rear of the drill bit. By emitting low-energy gamma rays and detecting their attenuation after passing through the rock and soil, it obtains the volume density of the rock and soil in real time during drilling, with a measurement accuracy of ±0.01 g / cm³. Density data is correlated or matched with the longitudinal wave velocity (Vp) and shear wave velocity (Vs) output by the velocity detection unit using a unified drilling depth coordinate, ensuring spatiotemporal consistency of velocity and density data within the same rock and soil layer.

[0059] In this embodiment, porosity calculation is based on the physical relationship between the elastic parameters and pore structure of the geotechnical medium. The porosity calculation unit incorporates a built-in porosity solution model, trained using a laboratory-established rock sample database (covering Vp, Vs, density, and measured porosity for common rock types). A multivariate regression algorithm or model is used to establish a nonlinear mapping relationship between Vp, Vs, density, and porosity. For special lithologies (such as fractured granite or porous sandstone), the porosity solution model dynamically loads a preset lithology correction coefficient table to adjust the porosity calculation weights, improving its applicability under different geological conditions. In an alternative embodiment, a physics-informed neural network (PINN) can be used instead of the multivariate regression model. This model embeds the physical relationship between geotechnical elastic parameters (Vp, Vs, density) and porosity, such as a theoretical model, into a neural network. Residual constraints are used to ensure that the model output conforms to known physical laws. Measured data is also used to train the nonlinear mapping relationship. In another alternative embodiment, a graph neural network model based on fracture structure representation, or a fissure graph neural network (GNN) model, is employed. By combining a graph neural network (GNN) with fracture structure representation, the ability to fit complex relationships can be enhanced. Specifically for fractured lithologies (such as fractured granite), this model abstracts the rock pore-fracture system into a graph structure, where nodes represent pores or fracture units and edges indicate connectivity. The GNN is then used to learn the mapping between node attributes (such as pore diameter and fracture length) and global elastic parameters, directly outputting porosity.

[0060] In this embodiment, the porosity solution model calculation component performs porosity solution in the following steps: in the data alignment step, the Vp, Vs and density data of the same depth point are matched according to the drilling depth coordinates; in the model calculation step, Vp, Vs and rock density are input into the porosity solution model, and the initial porosity value is calculated based on the statistical relationship; in the lithology correction step, the lithology type is inferred according to the real-time drilling parameters (such as drilling speed and torque), and the corresponding correction coefficient is called to optimize the porosity value; in the anomaly filtering step, if the calculated porosity exceeds the preset range (for example, 0-40%), it is marked as abnormal data and the review or verification process is triggered to eliminate errors caused by sensor noise or rock heterogeneity.

[0061] To ensure accuracy, the porosity calculation unit in this embodiment integrates laboratory calibration and field validation. In laboratory mode, rock samples with known porosity are placed in a simulated drilling environment. By comparing calculated values ​​with measured values, model parameters are calibrated to ensure a relative error of less than 5%. In field applications, the model correction coefficient is dynamically optimized by comparing porosity calculations for the same rock formation in adjacent boreholes with core laboratory test data. If the deviation between three consecutive depth points exceeds 8%, the model recalibration process is automatically triggered, updating the lithology-porosity correlation parameters.

[0062] In this embodiment, the calculated porosity parameters are stored in depth sequence and input into the early warning decision module simultaneously with stress response parameters, velocity variation parameters, and fracture evolution parameters. Porosity data is used to assess the permeability and strength degradation of the rock mass: areas with high porosity (e.g., >25%) indicate loose rock mass or the presence of invisible fractures, requiring priority support. Low porosity combined with high velocity gradients reflects the risk of stress concentration in dense rock formations. Porosity data is further incorporated into the calculation of a multi-parameter roof stability score. When the score exceeds the early warning threshold, a porosity distribution cloud map and risk zone location information are automatically generated to guide the optimization of underground engineering measures.

[0063] In this example, the porosity calculation results are output in JSON format, containing fields such as depth, porosity value, confidence level, and lithologic tag. This data is transmitted to the surface control system via the drill pipe's internal communication link and overlaid with the drilling trajectory data to generate a 3D porosity distribution model.

[0064] The beneficial effect of this technical solution is that by integrating the longitudinal and shear wave velocity data obtained by the wave velocity detection unit and the rock and soil density measured by the density sensor, the porosity parameters of the rock and soil body can be calculated in real time in situ, realizing the dynamic quantitative characterization of the pore structure; combining the physical correlation between wave velocity and density, the accuracy of porosity calculation is improved, and the lag and local defects of traditional sampling laboratory tests are compensated, providing key parameter support for evaluating the permeability and damage deterioration degree of rock and soil media, and collaborating with multi-source data such as stress and cracks to enhance the reliability and comprehensiveness of the roof rupture warning model.

[0065] In a further embodiment, each detection, monitoring or measurement unit of the geotechnical physical property detection module is synchronized with the power system or rotation system of the drilling equipment to achieve clock calibration, ensuring a one-to-one correspondence between the geotechnical physical property parameters and the drilling process data in terms of time axis and spatial depth.

[0066] Specifically, the geophysical property detection module integrates a synchronized clock architecture with the drilling equipment's power and rotation systems, including the stress response measurement unit, wave velocity detection unit, fracture evolution monitoring unit, and porosity calculation unit. The main control module, installed in the drilling rig's power compartment, utilizes a GPS-synchronized clock chip and a high-stability crystal oscillator dual-mode clock source to output a 1PPS (pulse per second) reference signal to each detection unit. The power system's motor encoder outputs real-time speed pulse signals, which are distributed via an RS-422 differential bus to the FPGA (Field-Programmable Gate Array) clock management module in each detection unit, achieving hardware-level clock synchronization between the drilling rotation and data acquisition.

[0067] Specifically, during drilling, the power system generates a synchronization calibration pulse every 10 rotations (corresponding to the drilling depth increment ΔH, converted from the drill pipe pitch). Upon receiving the pulse, each detection unit immediately performs the following operations: resets its internal clock counter to the current value of the power system clock to eliminate clock drift; records the current drilling depth coordinate (provided by the drill pipe length encoder); and inserts a synchronization marker into the acquired data stream, containing a timestamp (in microseconds) and depth value (in millimeters). The torque sensor data of the rotating system also serves as an auxiliary calibration source. When a sudden torque change is detected, an emergency synchronization event is triggered, forcing each unit to suspend acquisition and perform clock alignment.

[0068] Specifically, the third processing unit incorporates a built-in time-depth alignment engine, which performs spatiotemporal matching on the raw data uploaded by each detection unit. Based on the synchronized timestamps and depth values, the alignment engine constructs a time-depth interpolation model with drilling speed as the variable. For asynchronous data (e.g., depth deviations due to camera image processing delays), a sliding window interpolation method is used with a 10 ms window length to resample the data from each unit onto a unified spatiotemporal grid, ensuring that stress, velocity, fracture, and porosity parameters at the same time point precisely correspond to the same depth coordinate.

[0069] To ensure synchronization effectiveness, the system integrates dual-mode verification. In laboratory mode, the drill pipe idles at a fixed rotational speed (e.g., 60 rpm) and a constant feed rate (e.g., 0.1 m / min). Clock synchronization accuracy is verified by comparing the timestamp deviation (required to be less than 1 ms) and depth alignment error (required to be less than 2 mm) of the data from each detection unit. In field mode, if the depth deviation of a unit's synchronization mark exceeds 5 mm for three consecutive times, the unit's clock recalibration procedure is automatically triggered, and an alarm is issued through the human-machine interface to prompt maintenance. Historical synchronization status data is stored in log form to support fault backtracking analysis.

[0070] Specifically, through strict clock synchronization and depth alignment, stress response parameters, wave velocity change parameters, fracture evolution parameters, and porosity data are aligned one-to-one in time and spatial depth. The ground control system can overlay and display multiple parameters by depth coordinate. For example, at a certain depth point, it can simultaneously display a combination of data showing a sudden increase in axial stress, a sudden drop in longitudinal wave velocity, fracture density exceeding the threshold, and porosity anomalies, accurately indicating the concentrated damage area in the roof rock mass. Dynamic drilling parameters (such as rotation speed and drilling pressure) can also be analyzed in conjunction with geotechnical parameters. For example, when a high drilling speed is accompanied by a low wave velocity, it is determined that the drilling is in a loose and fractured zone, and the warning threshold is lowered in real time.

[0071] In a further embodiment, the geotechnical physical property parameters further include elastic modulus parameters, shear modulus parameters, and Poisson's ratio parameters of the geotechnical material below the stope roof. The geotechnical physical property detection module includes an elastic parameter calculation unit configured to calculate the elastic modulus parameters, shear modulus parameters, and Poisson's ratio parameters based on the longitudinal wave velocity and shear wave velocity measured by the wave velocity detection unit and the geotechnical density obtained by a density sensor associated with the drilling rig.

[0072] Specifically, the elastic parameter calculation unit is integrated into the data processing core of the geophysical property detection module and communicates in real time with the velocity detection unit and density sensor via a high-speed data bus. The longitudinal wave velocity (Vp) and shear wave velocity (Vs) data provided by the velocity detection unit, along with the geophysical density (ρ) values ​​collected by the density sensor, are synchronously transmitted to the elastic parameter calculation unit based on the drilling depth coordinate. The data interface utilizes a timestamp matching mechanism to ensure strict alignment of Vp, Vs, and ρ data at the same depth, avoiding calculation errors caused by transmission delays.

[0073] Specifically, the calculation of the elastic modulus (E) is based on the physical relationship between the longitudinal and shear wave velocities and density in elastic media. The elastic parameter calculation unit first calculates the shear modulus (G) based on the shear wave velocity Vs and the density ρ, using the formula G = ρ × Vs². Subsequently, the elastic modulus E is deduced using the relationship between the longitudinal wave velocity Vp and the shear modulus G, combined with the formula E = 2G(1 + ν) (ν is the Poisson's ratio). Since the Poisson's ratio ν is unknown, the calculation unit uses an iterative method: ν = 0.25 is initially assumed. After substituting it into the calculated E, the ν value is reversed using E = ρ × Vp² × (1 + ν)(1 - 2ν) / (1 - ν) until the difference between the two iterations is less than 0.001, outputting stable E and ν.

[0074] Specifically, the shear modulus (G) is calculated directly from the shear wave velocity Vs and density ρ using the formula G = ρ × Vs². The calculation unit applies a sliding average filter (with a window length of five adjacent depths) to the Vs data at the same depth point to eliminate velocity fluctuations caused by local rock mass heterogeneity and ensure the continuity and reliability of the G value. If the filtered Vs value deviates abnormally from the original data (for example, by more than 10%), the G value at that depth point is marked as low confidence and weighted in subsequent analysis.

[0075] Specifically, Poisson's ratio (ν) is calculated as the ratio of the elastic modulus E to the shear modulus G, using the formula ν = (E / (2G)) - 1. The calculation unit verifies the physical rationality of the ν value. If the calculated result exceeds the reasonable range for the geotechnical medium (for example, ν < 0.1 or ν > 0.4), the data review process is automatically triggered to recheck the input values ​​of Vp, Vs, and ρ for abnormalities or associated sensor failures. Passing the verification, ν values ​​are stored in depth sequence and, together with E and G, form the elastic parameter dataset.

[0076] To ensure accuracy, the elastic parameter calculation unit includes a built-in laboratory calibration database, covering the corresponding relationships between Vp, Vs, ρ and E, G, and ν for common lithologies (such as granite, sandstone, and shale). During on-site calculations, the current lithology type is inferred based on real-time drilling parameters (such as the drill bit vibration spectrum and drilling speed), and the corresponding calibration data is used to perform linear corrections on the calculated results. For example, if a high-frequency vibration signal is detected and the drilling speed drops sharply, it is determined to be a hard rock formation, and the granite calibration factor is used to apply a +5% correction to the E value.

[0077] Specifically, the calculated elastic parameters (E, G, ν) are integrated with stress response parameters, wave velocity variation parameters, fracture evolution parameters, and porosity data according to depth coordinates to form a multidimensional geotechnical mechanics database. For example, a combined analysis of the elastic modulus E and wave velocity Vp can identify areas of degraded rock mass elastic properties (E < 10 GPa and Vp < 3500 m / s); the combined parameters of Poisson's ratio ν and porosity are used to assess the potential for plastic deformation of the rock mass (ν > 0.3 and porosity > 20% are considered to indicate a prone to rheological formation). The data can also be rendered using the ground control system's 3D visualization module, generating color-coded distribution maps and trend curves of the elastic parameters along the borehole profile, enabling engineers to dynamically adjust roof support strategies.

[0078] Specifically, the elastic parameter calculation unit integrates an anomaly detection mechanism: when the E, G, and ν values ​​at three consecutive depth points exceed preset lithologic thresholds, an audible and visual alarm is triggered, and drilling is suspended. Simultaneously, the system's self-check module automatically verifies the data integrity and clock synchronization status of the velocity detection unit and density sensor, generating a fault diagnosis report. For example, if the G value is abnormal while Vs and ρ are normal, it is determined that the shear wave signal is being interfered with by the borehole fluid, prompting the activation of the secondary filtering algorithm for the velocity data.

[0079] Specifically, the elastic parameter dataset is output in a structured format, containing the following fields: depth, E (in GPa), G (in GPa), ν, confidence level (0-1), and correction flags. This data is uploaded in real time via downhole optical fiber to the surface control system or data processing module, where it participates in the weighted calculation of the roof stability scoring model. For example, the elastic modulus E, a key indicator of rock mass compressive strength, is weighted at 30%. Together with fracture density (25%), velocity gradient (20%), porosity (15%), and stress state (10%), it forms a comprehensive evaluation system that outputs a fracture risk level or recommended actions.

[0080] In a further embodiment, the sensor array in the drilling data acquisition module includes: at least one drilling pressure sensor for real-time monitoring of the drilling pressure applied by the drill tool to the rock formation during drilling; at least one torque sensor for real-time monitoring of the torque of the drill tool during rotation during drilling; at least one drilling speed sensor for real-time monitoring of the drilling speed of the drill tool during drilling; the sensor array establishes a communication connection with the data processing module via a wireless transmission module or a wired cable to achieve real-time synchronization of drilling process data.

[0081] In the drilling data acquisition module, at least one WOB sensor is installed at a suitable location on the drill string, ensuring it can accurately sense the WOB applied by the drill string to the rock formation. During the drilling process, the drill string continuously applies pressure to the rock formation to achieve drilling, and the WOB sensor monitors this pressure in real time. It converts the sensed WOB into an electrical signal that contains real-time WOB information. For example, when the drill string encounters harder rock, the WOB increases, and the electrical signal output by the WOB sensor changes accordingly, reflecting the real-time WOB status.

[0082] At least one torque sensor is installed on the rotating part of the drill string to monitor the torque generated by the drill string in real time. During the drilling process, the drill string rotates to break up the rock, generating torque. The torque sensor accurately captures the magnitude of this torque. It converts the sensed torque into an electrical signal, which provides information about the torque experienced by the drill string during rotation. For example, as the rock hardness increases, the resistance to the drill string's rotation increases, and the torque also increases. This change is reflected in the electrical signal output by the torque sensor.

[0083] At least one drilling speed sensor is installed on the drill string to monitor its penetration rate in real time. During drilling, the drill string moves in a specific direction, and the drilling speed sensor accurately measures its speed. It converts the penetration rate into an electrical signal, providing the system with real-time information on the drilling speed. For example, as the drill string encounters rock formations of varying hardness, the penetration rate changes, and the electrical signal output by the drilling speed sensor also changes accordingly, reflecting the real-time penetration rate.

[0084] The electrical signals collected by the weight-on-bit sensors, torque sensors, and drilling speed sensors in the sensor array need to be transmitted to the data processing module for further analysis and processing. To achieve this, the sensor array establishes a communication connection with the data processing module via a wireless transmission module or a wired cable.

[0085] In a further embodiment, the system further includes a drilling optimization module for dynamically adjusting drilling control parameters according to the geophysical property parameters and the drilling process data.

[0086] Specifically, the drilling optimization module consists of a data interface unit, a decision algorithm library, and a control command output unit. The data interface unit receives real-time geophysical property parameters (stress response parameters, wave velocity variation parameters, fracture evolution parameters, porosity, elastic modulus, etc.) and drilling process data (weight on bit, torque, and drilling speed) via a fiber optic bus. It performs spatiotemporal alignment based on depth coordinates to construct a multidimensional data fusion matrix. The decision algorithm library integrates a rule engine and machine learning models, and the control command output unit connects to the drilling rig's PLC control system via the Modbus protocol to enable dynamic adjustment of drilling parameters.

[0087] Specifically, the module includes a library of preset adjustment rules based on expert experience. For example, if the maximum principal stress at a certain depth exceeds 20 MPa and the fracture density exceeds 15 fractures / meter, it is identified as a high-risk fracture zone. The drill pressure is automatically reduced to 70% of the rated value, the drilling speed is limited to 0.3 m / h, and the flushing fluid flow rate is increased to 150 L / min to stabilize the borehole wall. If the porosity exceeds 25% and the shear wave velocity gradient exceeds 400 m / s / m, it is identified as a loose rock formation and the "low drill pressure-high rotation speed" mode (8 tons of drill pressure, 120 rpm) is triggered to reduce the risk of drill bit sticking. If the elastic modulus E is less than 8 GPa and the Poisson's ratio ν is greater than 0.35, it is identified as a soft rock plastic zone and an intermittent drilling strategy is initiated (drilling for 30 seconds followed by a 5-second pause) to prevent overheating and wear of the drill bit.

[0088] Specifically, machine learning models (such as deep Q networks) in the decision-making algorithm library continuously optimize control strategies based on historical drilling data. The model input layer includes current geotechnical parameters, drilling parameters, and data trends from the previous 10 depth points. The output layer provides adjustment recommendations for drilling pressure, rotational speed, and flushing fluid flow rate. The model evaluates the control effect using a reward function: if the drilling speed fluctuation coefficient decreases and the drill bit vibration amplitude decreases after adjustment, the weight of the strategy is increased; if an abnormal jump in drilling pressure or torque exceeds the limit occurs, it is marked as negative feedback and the strategy library is updated. After each hole is drilled, the model automatically generates an optimization report and pushes it to the ground control system for manual review and strategy fine-tuning.

[0089] Specifically, the control command output unit converts decision results into control signals recognizable by the drilling rig. For example, the WOB adjustment command adjusts the hydraulic system's oil supply pressure via a proportional valve with an accuracy of ±0.5 tons; the speed adjustment command is implemented by a frequency converter driving the motor with a response time of less than 2 seconds; and the flushing fluid flow rate is controlled by an electric control valve with a flow rate tolerance of ±3%. Simultaneously, the module monitors the adjusted drilling data (such as the deviation between the adjusted WOB and the target value and changes in the vibration sensor spectrum) in real time, forming a closed-loop feedback loop. If the actual WOB deviates from the target value by more than 15% for 10 seconds, an execution anomaly is detected, triggering a backup control strategy (such as switching to manual remote control mode).

[0090] Specifically, the ground control interface offers three control modes: automatic, semi-automatic, and manual. In automatic mode, drilling parameters are dynamically adjusted entirely by the optimization module. In semi-automatic mode, engineers can preset a parameter adjustment range (e.g., 10-25 tons of weight on bit), and the module autonomously optimizes within this range. In manual mode, recommended optimization parameters are displayed in real time and executed upon manual confirmation. If the system detects a sudden change in geotechnical parameters outside the model's training range (e.g., longitudinal wave velocity Vp < 2000 m / s) or drill string vibration acceleration exceeding 5g, drilling is automatically suspended and an audible and visual alarm is sounded, prompting the activation of emergency response plans (e.g., withdrawing the drill pipe and implementing temporary support). In a further embodiment, the data processing module includes: a data preprocessing unit, used to preprocess the geotechnical physical property parameters and the drilling process data to obtain preprocessed geotechnical physical property parameters and drilling process data; an algorithm calculation unit, used to input the preprocessed geotechnical physical property parameters and drilling process data into the hard rock damage evolution algorithm for calculation to obtain an intermediate calculation result associated with the damage state of the mine roof; a risk assessment unit, used to determine the fracture risk of the mine roof based on the intermediate calculation result.

[0091] In this embodiment, the data preprocessing unit receives parameters such as stress, velocity, cracks, porosity, and drilling process data (weight on bit, torque, and speed) from the geophysical property detection module and performs multi-stage preprocessing. First, data cleaning is performed: missing segments caused by abnormal sensor power outages or communication interruptions are removed and filled using linear interpolation of adjacent depth data. Noise data (such as high-frequency glitches in the stress signal) is denoised using a median filter combined with a sliding average filter (with a window length of 5 depth points). Next, normalization is performed: each parameter is dimensionally normalized to the range of 0-1. For example, stress values ​​are divided by the maximum sensor range, and velocity values ​​are scaled according to the laboratory calibration range. Finally, spatiotemporal alignment is performed: based on the drilling depth coordinate, data with different sampling frequencies (e.g., 200 Hz stress data and 10 Hz velocity data) are resampled to a uniform depth interval (0.1 meter / point) using a sliding window interpolation method to ensure strict spatial alignment of the multi-source data. The preprocessed data is stored in a structured array by depth sequence for subsequent algorithm invocation.

[0092] In this embodiment, the algorithm calculation unit incorporates a hard rock damage evolution algorithm, which can be used as a time series prediction model based on multi-source data fusion. The model input layer includes preprocessed geotechnical parameters and drilling data, with time-dimensional features constructed by depth sequence. The algorithm first extracts time series features using an LSTM (Long Short-Term Memory) network to capture the dynamic evolution trends of damage parameters (e.g., the accumulation rate of crack density with depth and the abrupt change point of the velocity gradient). Subsequently, a random forest model is combined to perform nonlinear modeling of multi-parameter interactions, for example, analyzing the contribution weight of the combination of high drilling pressure and low elastic modulus to the damage factor. Intermediate calculation results include: damage factor D (on a 0-1 scale, where 0 indicates intact rock mass and 1 indicates complete damage); damage accumulation rate ΔD / ΔH (damage increment per unit depth); and critical damage threshold D_c (an empirical value derived from inversion of historical roof failure cases). Model training data is derived from geotechnical parameters, drilling data, and corresponding roof failure records from historical boreholes. Model parameters are dynamically updated quarterly with new field data.

[0093] In this embodiment, the risk assessment unit performs multi-dimensional risk quantification based on the intermediate calculation results (D, ΔD / ΔH, and D_c). First, the risk level is determined based on the current value and historical trend of the damage factor D: if D ≥ 0.8D_c and ΔD / ΔH > 0.05 / m, the risk is marked as high; if 0.5D_c ≤ D < 0.8D_c and ΔD / ΔH > 0.03 / m, the risk is marked as medium; all other risk levels are low. Second, a fuzzy comprehensive evaluation method is used to integrate the contributions of multiple parameters. For example, when the fracture fractal dimension is greater than 1.6 and the porosity is greater than 20%, the risk level is automatically increased by one level. The final output is a fracture risk index R (0-100), calculated using the formula: R = αD / D_c + β(ΔD / ΔH) / 0.05 + γΣ (other parameter weights). α, β, and γ are weight coefficients dynamically adjusted based on geological conditions. Other parameters refer to key parameters introduced in the model, in addition to the damage factor (D / D_c) and damage rate (ΔD / ΔH). Σ (other parameter weights) represents the sum of the weighted contributions of these key parameters. Risk assessment results trigger a graded warning: a red warning (immediate evacuation and support) is activated when R ≥ 75; a yellow warning (limited drilling speed and monitoring) is activated when 50 ≤ R < 75; and a green safety status is activated when R < 50. Warning signals are simultaneously issued via underground audible and visual alarms and the surface monitoring interface. The warning signal is linked to a 3D model of the borehole location, highlighting the spatial coordinates of the risk area and detailed damage parameters.

[0094] In this embodiment, the ground control system provides an interactive dashboard that dynamically displays preprocessed data, intermediate calculation results, and risk assessment results. For example, a curve of the damage factor D along the borehole depth is superimposed with the velocity gradient and fracture density curves to help locate areas of concentrated damage. The risk index R is mapped to the 3D tunnel model as a color block diagram, with red blocks indicating areas with a high risk of fracture. Engineers can manually adjust risk assessment weights (for example, temporarily increasing the contribution factor of porosity) or flag false positives to optimize the model. The system records all manual interventions for subsequent model training iterations.

[0095] The advantages of this technical solution are that the data preprocessing unit cleans, normalizes and aligns multi-source geotechnical parameters and drilling data in time and space, ensuring the quality of input data for subsequent analysis; the algorithm calculation unit dynamically integrates multi-dimensional parameters based on the hard rock damage evolution algorithm, accurately solves the roof damage status, and realizes quantitative characterization of the damage degree; the risk assessment unit combines the damage status with engineering experience thresholds to generate graded warning signals, which can reflect the spatiotemporal evolution trend of the mine roof fracture risk in real time, thereby improving the timeliness and accuracy of the warning.

[0096] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0097] like Figure 7 As shown, an embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the method for intelligent early warning of stope roof fracture based on hard rock damage evolution as described above is implemented, which includes the following steps:

[0098] S10: obtaining geotechnical property parameters of the rock and soil below the stope roof based on measurement while drilling during the drilling process, wherein the geotechnical property parameters include stress response parameters, wave velocity variation parameters, crack evolution parameters, and porosity parameters of the rock and soil;

[0099] S20: Acquire drilling process data including bit weight, torque, and drilling speed during the drilling process, wherein the drilling data acquisition module includes a sensor array connected to the drilling equipment, and is used to monitor and record the drilling process data in real time;

[0100] S30: obtaining an analysis result associated with the damage state of the stope roof based on the geotechnical physical property parameters and the drilling process data in combination with a hard rock damage evolution algorithm, and determining the fracture risk of the stope roof based on the analysis result.

[0101] If the integrated modules / units are implemented as software functional units and sold or used as standalone products, they can be stored in a computer-readable storage medium. Based on this understanding, the present invention can also implement all or part of the process steps in the above-mentioned method embodiments by using a computer program to instruct the relevant hardware. The computer program can be stored in a computer-readable storage medium. When executed by a processor, the computer program can implement the steps of each of the above-mentioned method embodiments. The computer program includes computer program code, which can be in source code form, object code form, executable file, or some intermediate form. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium. Of course, there are other types of readable storage media, such as quantum memory and graphene memory. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practices in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practices, computer-readable media do not include electrical carrier signals and telecommunication signals.

[0102] The present invention also provides an electronic device. The electronic device in an embodiment of the present invention includes: one or more processors; and a storage device for storing one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the intelligent early warning method for stope roof fracture based on hard rock damage evolution provided by the present invention.

[0103] Reference below Figure 8 , which shows a schematic structural diagram of a computer system 800 of an electronic device suitable for implementing an embodiment of the present invention. Figure 8 The electronic device shown is only an example and should not limit the functions and scope of use of the embodiments of the present invention.

[0104] like Figure 8 As shown, computer system 800 includes a central processing unit (CPU) 801, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 802 or a program loaded from a storage unit 808 into a random access memory (RAM) 803. Various programs and data required for the operation of computer system 800 are also stored in RAM 803. CPU 801, ROM 802, and RAM 803 are connected to each other via a bus 804. An input / output (I / O) interface 805 is also connected to bus 804.

[0105] The following components are connected to the I / O interface 805: an input section 806 including a keyboard, mouse, and the like; an output section 807 including devices such as a cathode ray tube (CRT), a liquid crystal display (LCD), and speakers; a storage section 808 including a hard disk; and a communication section 809 including a network interface card such as a LAN card or a modem. The communication section 809 performs communication processing via a network such as the Internet. A drive 810 is also connected to the I / O interface 805 as needed. Removable media 811, such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory, is installed in the drive 810 as needed, so that computer programs read from the removable media can be installed in the storage section 808 as needed.

[0106] In particular, according to embodiments disclosed herein, the processes described in the main step diagrams above can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for executing the methods shown in the main step diagrams. In the above embodiments, the computer program can be downloaded and installed from a network via the communication section 809 and / or installed from removable media 811. When the computer program is executed by the central processing unit 801, the above-described functions defined in the system of the present invention are performed.

[0107] It should be noted that the computer-readable medium described in the present invention may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. Computer-readable storage media may include, for example, but not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or components, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to, an electrical connection having one or more conductors, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device, or component. In the present invention, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such a propagated data signal may take a variety of forms, including, but not limited to, electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device. Program code embodied on a computer-readable medium may be transmitted using any suitable medium, including but not limited to wireless, wireline, optical cable, RF, or any suitable combination thereof.

[0108] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a module, program segment or a part of code, and the part of the above-mentioned module, program segment or code comprises one or more executable instructions for realizing the logical function of the specification. It should also be noted that in some alternative implementations, the function marked in the box can also occur in a different order than that marked in the accompanying drawings.

[0109] The above specific embodiments do not limit the scope of protection of the present invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may occur depending on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. An intelligent early warning system for stope roof fracture based on hard rock damage evolution, characterized by: include: A geotechnical property detection module is used to obtain geotechnical property parameters of the rock and soil below the stope roof based on measurement while drilling during drilling. The geotechnical property parameters include stress response parameters, wave velocity change parameters, crack evolution parameters, and porosity parameters of the rock and soil; A drilling data acquisition module, configured to acquire drilling process data including weight on bit, torque, and drilling speed during the drilling process. The drilling data acquisition module includes a sensor array connected to the drilling equipment for real-time monitoring and recording of the drilling process data. a data processing module, configured to obtain, based on the geophysical property parameters and the drilling process data, an analysis result associated with the damage state of the stope roof in combination with a hard rock damage evolution algorithm, and determine the fracture risk of the stope roof based on the analysis result; The geotechnical physical property detection module includes a stress response measurement unit and a first processing unit; The stress response measurement unit is configured with a stress sensor array arranged along the drilling direction, the stress sensor array is connected to the drill pipe of the drilling equipment, and is used to collect original signals corresponding to the hoop stress, axial stress and radial stress at different drilling depths; The first processing unit is used to calculate the stress response parameters of the rock and soil through a stress tensor solution model based on the original signals corresponding to the hoop stress, axial stress and radial stress at different drilling depths.

2. The intelligent early warning system for stope roof fracture based on hard rock damage evolution according to claim 1 is characterized in that: The geotechnical physical property detection module further includes a wave velocity detection unit and a second processing unit; The wave velocity detection unit includes an acoustic wave transmitting probe installed in the drill bit and an acoustic wave receiving array arranged in the middle of the drill pipe. The acoustic wave transmitting probe is used to transmit pulsed acoustic waves, and the acoustic wave receiving array is used to collect reflected signals and transmitted signals from different interfaces and generate initial detection data. The initial detection data includes signal arrival time, signal amplitude, and phase information; wherein the signal amplitude refers to the intensity change of the reflected signal and the transmitted signal, and the phase information refers to the phase difference between the reflected signal and the transmitted signal; The second processing unit is configured to calculate, based on the initial detection data and the corresponding drilling depth coordinates, a correspondence between the propagation time and the propagation distance of the acoustic wave signal at different depths based on a time-distance curve fitting algorithm, and calculate the longitudinal wave velocity and the shear wave velocity of the geotechnical medium based on the correspondence and the physical properties of the geotechnical medium; The second processing unit is further configured to obtain a velocity variation parameter by performing difference calculation and gradient analysis on the longitudinal wave velocity and the shear wave velocity at adjacent depths.

3. The intelligent early warning system for stope roof fracture based on hard rock damage evolution according to claim 2 is characterized in that: The rock and soil physical property detection module also includes a crack evolution monitoring unit and a third processing unit; The fracture evolution monitoring unit is used to obtain hole wall images through a camera integrated into the side wall of the drill pipe; The third processing unit is used to use a digital image recognition algorithm to perform edge detection and contour tracing on the cracks in the hole wall image, quantify the spatial morphological parameters of the cracks and the connection relationship between the crack endpoints to obtain the connectivity of the cracks, calculate the crack density based on the number of cracks within the unit hole wall length, and calculate the crack fractal dimension based on the fractal box dimension algorithm of the crack geometric distribution and the connectivity; the crack evolution parameters include the crack density and the crack fractal dimension.

4. The intelligent early warning system for stope roof fracture based on hard rock damage evolution according to claim 3 is characterized in that: The rock and soil physical property detection module also includes a porosity calculation unit for calculating the porosity parameter by combining the longitudinal wave velocity and shear wave velocity measured by the wave velocity detection unit and the rock and soil density obtained by the density sensor of the drilling rig.

5. The intelligent early warning system for stope roof fracture based on hard rock damage evolution according to any one of claims 1 to 4, characterized in that: The rock and soil physical property detection module and the power system or rotation system of the drilling equipment achieve synchronous clock calibration.

6. The intelligent early warning system for stope roof fracture based on hard rock damage evolution according to claim 2 is characterized in that: The rock and soil physical property parameters also include the elastic modulus parameters, shear modulus parameters and Poisson's ratio parameters of the rock and soil below the mine roof; the rock and soil physical property detection module also includes: an elastic parameter calculation unit, which is used to calculate the elastic modulus parameters, the shear modulus parameters and the Poisson's ratio parameters based on the longitudinal wave velocity and shear wave velocity measured by the wave velocity detection unit, and the rock and soil density obtained by the drilling rig's matching density sensor.

7. The intelligent early warning system for stope roof fracture based on hard rock damage evolution according to claim 1 is characterized in that: The sensor array in the drilling data acquisition module includes: At least one weight-on-bit sensor for monitoring the weight-on-bit applied by the drill tool to the rock formation during drilling in real time; At least one torque sensor for real-time monitoring of the torque of the drill bit during rotation during drilling; At least one drilling speed sensor for real-time monitoring of the drilling speed of the drilling tool during drilling; The sensor array establishes a communication connection with the data processing module via a wireless transmission module or a wired cable to achieve real-time synchronization of drilling process data.

8. The intelligent early warning system for stope roof fracture based on hard rock damage evolution according to claim 1 is characterized in that: It also includes a drilling optimization module for dynamically adjusting drilling control parameters according to the rock and soil physical property parameters and the drilling process data.

9. The intelligent early warning system for stope roof fracture based on hard rock damage evolution according to claim 1 is characterized in that: The data processing module includes: a data preprocessing unit, configured to preprocess the geotechnical physical property parameters and the drilling process data to obtain preprocessed geotechnical physical property parameters and drilling process data; an algorithm calculation unit, configured to input the pre-processed geotechnical physical property parameters and drilling process data into a hard rock damage evolution algorithm for calculation, and obtain an intermediate calculation result associated with the damage state of the stope roof; The risk assessment unit is used to determine the breakage risk of the stope roof according to the intermediate calculation result.

Images