Method for observing drilling hole and peeping blasting crack in three-dimensional mode through miniature machine

Abstract

The invention discloses a miniature machine three-dimensional observation drilling peeping blasting crack method, and belongs to the technical field of drilling observation technology and blasting engineering detection.The miniature machine three-dimensional observation drilling peeping blasting crack method includes the steps that a miniature three-dimensional scanning machine is designed, matched with a look-around camera array and a laser scanning unit and fed into a to-be-detected drilling hole through a flexible guiding device, and autonomous navigation and data fusion technologies are combined; and high-precision generation of the drilling three-dimensional digital rock core diagram and the trajectory diagram is realized. Through comparison of data observed before and after blasting, crack parameters are quantitatively extracted, optimal blasting parameters are reversely deduced in combination with detonation pressure, a charging structure and geological conditions, and a closed-loop optimization system is formed. According to the method, the observation efficiency is greatly improved, the view blind area is overcome, the method adapts to the complex drilling environment, key technical support is provided for intelligent blasting and digital mine construction, and the method has remarkable practicability and innovativeness.

Description

Technical Field

[0001] This invention belongs to the field of borehole observation technology and blasting engineering detection technology, specifically relating to a method for three-dimensional observation of boreholes and inspection of blasting cracks using a micromachine. Background Technology

[0002] Currently, borehole inspection technology plays a crucial role in engineering and science, particularly in assessing and monitoring borehole quality and crack surface conditions. Typically, this is achieved using borehole inspection instruments. These instruments are ingeniously designed, usually consisting of a connecting rod equipped with a camera, made of a rigid, solid material to ensure accurate observation inside the borehole.

[0003] However, the situation is slightly different for directional drilling. Directional drilling is a technically complex and relatively precise drilling method, in which the borehole may be at a certain angle. This design means that the line of sight through the borehole is not a straight line, but rather has a certain angle, which increases the difficulty of borehole inspection. Summary of the Invention

[0004] The purpose of this invention is to provide a method for three-dimensional observation of boreholes and blasting cracks using a micromachine, so as to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for three-dimensional observation of boreholes using a micromachine to inspect blasting cracks, comprising the following steps:

[0006] S1. Design and manufacture a miniature 3D scanning machine. The main body length of the 3D scanning machine is less than the diameter of the borehole to be measured. It is equipped with a surround-view camera array, a laser scanning unit, an inertial measurement unit (IMU), and an autonomous navigation system. It is also waterproof, dustproof, and vibration resistant, and its working temperature is adaptable to an environment of -10℃ to 60℃.

[0007] S2. The three-dimensional scanning machine is sent into the borehole to be tested by a flexible guide device, and the autonomous navigation system is started to control the three-dimensional scanning machine to move at a constant speed along the borehole axis.

[0008] S3. During the movement of the three-dimensional scanning machine, the surround-view camera array continuously captures panoramic images of the hole wall, the laser scanning unit synchronously collects the three-dimensional coordinate point cloud of the hole wall surface, and the inertial measurement unit corrects the attitude and depth positioning of the three-dimensional scanning machine in real time.

[0009] S4. The ground terminal receives images and point cloud data, and generates high-precision three-dimensional digital core images and borehole trajectory images through data fusion processing.

[0010] S5. Before and after blasting, perform the observation procedures from S2 to S4 on the same borehole to be tested to obtain the three-dimensional digital core map, trajectory map and corresponding raw data of the two observations.

[0011] S6. Compare the observation data before and after the blasting to extract the location, number, opening, extension direction and distribution density of the newly added cracks;

[0012] S7. Based on the detonation pressure, charge structure and geological conditions of the blasting project, analyze the fracture development law and deduce the optimal hole radius and blasting parameter combination.

[0013] S8. Feed back the optimal combination of blasting parameters to the next cycle of blasting design.

[0014] In one specific implementation, the surround-view camera array consists of 4 to 8 CMOS sensors, which are evenly distributed in a circle, and the resolution of a single CMOS sensor is not less than 1080P.

[0015] In the above implementation process, the parameters of the surround-view camera array need to be calibrated first. The miniature 3D scanner is placed in front of the 360° panoramic calibration board, and the exposure parameters and white balance of each CMOS sensor are adjusted to ensure that the imaging brightness and color of adjacent sensors are consistent. Then, images of the calibration board are acquired, and the stitching error is verified through an image stitching algorithm to ensure that the stitching gap of the 360° panoramic image is less than 1 pixel, eliminating visual blind spots. After the machine enters the drilling hole, each CMOS sensor synchronously captures images at a frequency of 30 frames per second. Every 5 sets of panoramic images are acquired, the system automatically matches and verifies the texture features of the image edges. If stitching misalignment occurs (deviation exceeding 2 pixels), real-time distortion correction is triggered, and the image is geometrically corrected using pre-stored drilling diameter parameters. For areas with strong reflections on the hole wall (such as dry rock surfaces), the system automatically reduces the exposure intensity of the corresponding CMOS sensor to avoid overexposure and loss of crack details, ensuring the clarity and integrity of the hole wall image.

[0016] In one specific implementation, the laser scanning unit is a single-line or multi-line rotating lidar with a scanning frequency of not less than 20Hz and a ranging error of not more than ±2mm.

[0017] During the above implementation process, the laser scanning unit needs to be zero-point calibrated beforehand. The miniature 3D scanning machine is placed in a standard calibration fixture, and the laser radar is activated to scan a calibration block with known dimensions on the inner wall of the fixture. The scanned data is compared with the actual dimensions of the calibration block, and the ranging error is corrected to within ±0.5mm to ensure the accuracy of subsequent hole wall point cloud acquisition. After the machine enters the borehole, the laser scanning unit continuously outputs point cloud data at a frequency of 20Hz, while simultaneously synchronizing with the image frames of the surround-view camera array. Every 10 sets of point cloud data are acquired, the coordinates are corrected once using the attitude data from the inertial measurement unit to avoid point cloud offset caused by slight machine shaking. If abnormal data with a ranging error exceeding ±2mm occurs during the scanning process (such as the presence of protruding debris obstructing the hole wall), the system will automatically trigger a rescanning mechanism, controlling the machine to stay at the current position for 0.5 seconds and repeat the scan 3 times. The average of the three data is taken as the valid data to ensure the accuracy of the 3D morphology reconstruction of the hole wall.

[0018] In one specific implementation, data transmission in S4 can be either wired or wireless. Wireless transmission uses 5G or LoRa communication protocols, supports remote control and real-time data analysis, and the data transmission latency does not exceed 50ms.

[0019] In the above implementation process, the wired transmission uses highly shielded coaxial cable or optical fiber to ensure the stability of data transmission in complex electromagnetic environments, with a transmission rate of no less than 1Gbps to meet the needs of large-scale real-time transmission of raw data. To ensure data integrity, the transmission protocol incorporates a triple verification mechanism, including CRC cyclic redundancy check, data packet sequence number check, and timestamp synchronization check, keeping the erroneous data packet retransmission rate below 0.1%. Under extreme conditions, the system automatically switches to the backup transmission channel, with a primary / backup channel switching time of less than 100ms, ensuring uninterrupted observation processes.

[0020] In one specific implementation, the flexible guide device is made of high-strength nylon braided material, and its diameter can be adaptively adjusted within the range of 8 to 50 mm to adapt to boreholes of different diameters. It can also guide the three-dimensional scanning machine through borehole sections with a bending angle of no more than 15°.

[0021] In the above implementation process, the surface of the flexible guide device undergoes special treatment, exhibiting excellent wear resistance and corrosion resistance, enabling it to operate stably for extended periods in complex geological environments. Simultaneously, its internal elastic support structure ensures sufficient rigidity during adaptive diameter adjustment, preventing damage due to excessive bending.

[0022] In one specific implementation, the autonomous navigation system has a built-in preset trajectory planning module and a real-time obstacle avoidance module. Through the coordinated feedback of the inertial measurement unit and the laser scanning unit, the moving speed of the three-dimensional scanning machine is adjustable within the range of 0.05 to 0.2 m / s, and the moving deviation does not exceed ±3 mm / m.

[0023] During the aforementioned implementation process, the autonomous navigation system can adjust its path planning in real time according to environmental changes, ensuring stable operation of the 3D scanning machine in complex terrain. When encountering obstacles, the real-time obstacle avoidance module is activated immediately, quickly calculating the optimal detour path through joint feedback from the inertial measurement unit and the laser scanning unit, with the entire response time not exceeding 0.5 seconds. Simultaneously, the preset trajectory planning module automatically adjusts the scanning machine's speed and direction based on the curvature and diameter changes of the borehole, ensuring the integrity and accuracy of the scanning data.

[0024] In one specific implementation, the data fusion processing in S4 adopts an image-point cloud registration algorithm based on Kalman filtering, which fuses the texture information of the two-dimensional image with the spatial location information of the three-dimensional point cloud, and the generated three-dimensional digital core image of the borehole has a resolution of not less than 0.1 mm / pixel.

[0025] In the above implementation process, the algorithm can effectively reduce noise interference, improve data accuracy, and ensure that the generated 3D digital core images of the borehole truly reflect the internal structural characteristics of the borehole. At the same time, the algorithm has high operating efficiency, capable of completing large-scale data processing in a short time, meeting the needs of real-time detection.

[0026] In one specific implementation scheme, S6 uses a deep learning semantic segmentation model to compare the observation data before and after the blasting, automatically identify newly added crack areas, and quantify the crack opening measurement accuracy to be no less than ±0.05mm and the extension direction measurement accuracy to be no less than ±2°.

[0027] In the aforementioned implementation process, this deep learning semantic segmentation model was trained on a large amount of labeled data, enabling it to accurately distinguish between cracked and non-cracked areas in images before and after blasting, effectively avoiding errors and omissions that may occur with manual identification. Simultaneously, the model possesses strong adaptive capabilities, adapting to crack characteristics under different geological conditions, ensuring high-precision crack identification and measurement in various scenarios.

[0028] In one specific implementation scheme, the geological conditions combined in S7 include rock mass compressive strength, rock mass integrity coefficient, rock stratum dip angle and joint development density. By establishing a multivariate regression model of fracture development parameters with detonation pressure, charge structure and geological conditions, the optimal combination of blasting parameters is deduced.

[0029] In the above implementation process, the training data for the multiple regression model came from historical blasting engineering records. After data cleaning and feature engineering, the integrity and accuracy of the input parameters were ensured. The model used the least squares method for parameter estimation and optimized the model structure through cross-validation. The final determined regression equation had a goodness of fit of no less than 0.85, and the prediction error was controlled within ±5%.

[0030] In one specific implementation, the 3D scanning machine is also equipped with a removable battery pack with a single-charge runtime of no less than 4 hours, supports hot-swappable replacement, and has a built-in data storage module that can locally store no less than 100GB of observation data.

[0031] To ensure the equipment's continuous operation in complex geological environments during the aforementioned implementation process, the 3D scanning machine's battery pack uses high-density lithium-ion cells, supporting a wide operating temperature range of -20℃ to 50℃. It is also equipped with an intelligent power management system that monitors remaining power and optimizes power allocation in real time. The data storage module uses a solid-state drive architecture with a ruggedized design, achieving IP68 protection rating, effectively resisting vibration, moisture, and dust interference. It also supports encrypted transmission, allowing observation data to be synchronized to a cloud server in real time via Wi-Fi or a 4G / 5G module.

[0032] Compared with the prior art, the beneficial effects of the present invention are:

[0033] 1. This invention eliminates the need for the segmented installation and disassembly of rigid connecting rods. Observation is completed directly through a flexible guidance device and an autonomous navigation machine, reducing preparation time by more than 50% and significantly improving construction efficiency. Simultaneously, the flexible guidance device possesses excellent bending adaptability, easily traversing tortuous sections and narrow areas within the borehole, ensuring the autonomous navigation machine smoothly reaches the target observation position, greatly enhancing the equipment's operational capabilities under complex geological conditions. The autonomous navigation machine is equipped with a high-precision three-dimensional scanning sensor, capable of performing comprehensive, high-precision scanning imaging of blasting cracks, acquiring detailed geometric parameters such as length, width, depth, and orientation, providing accurate and reliable data support for subsequent engineering analysis and safety assessment. Furthermore, this equipment boasts advantages such as high automation and ease of operation, reducing manual intervention, lowering labor intensity, and further improving the safety and reliability of construction.

[0034] 2. This invention employs a 360° surround-view camera array composed of 4 to 8 CMOS sensors evenly distributed in a circle, achieving 360° imaging without blind spots and completely solving the problem of blind spots in traditional equipment, ensuring the completeness of crack information capture. Because each CMOS sensor in the surround-view camera array has high resolution and fast response characteristics, it can capture subtle changes in blasted cracks in a short time, providing rich and accurate image information for subsequent data analysis. Moreover, this array design also has strong anti-interference capabilities, allowing it to operate stably even in complex construction environments with interference factors such as dust and vibration, ensuring image clarity and accuracy, and greatly improving the applicability and reliability of the equipment under harsh conditions.

[0035] 3. This invention analyzes data before and after blasting to deduce the optimal blasting parameters and feeds them back to subsequent blasting designs, thereby achieving dynamic optimization of blasting parameters, promoting the development of intelligent blasting, and providing key sensing technology support for the construction of digital mines and smart tunnels. Attached Figure Description

[0036] Figure 1 This is the vertical crack observation structure of the present invention;

[0037] Figure 2 This is a cross-sectional three-dimensional structural schematic diagram of the present invention, showing a vertical crack detection diagram. Detailed Implementation

[0038] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0039] Please see Figure 1 and Figure 2 This invention provides a method for three-dimensional observation of boreholes and inspection of blasting cracks using a micromachine, comprising the following steps:

[0040] S1. Design and manufacture a miniature 3D scanning machine. The main body length of the 3D scanning machine is less than the diameter of the borehole to be measured. It is equipped with a surround-view camera array, a laser scanning unit, an inertial measurement unit (IMU) and an autonomous navigation system. It is also waterproof, dustproof and vibration resistant, and its working temperature is adaptable to an environment of -10℃ to 60℃.

[0041] S2. Send the 3D scanning machine into the borehole to be tested through the flexible guide device, start the autonomous navigation system, and control the 3D scanning machine to move at a constant speed along the borehole axis.

[0042] S3. During the movement of the 3D scanning machine, the surround-view camera array continuously captures panoramic images of the hole wall, the laser scanning unit synchronously collects the 3D coordinate point cloud of the hole wall surface, and the inertial measurement unit corrects the attitude and depth positioning of the 3D scanning machine in real time.

[0043] S4. The ground terminal receives images and point cloud data, and generates high-precision three-dimensional digital core images and borehole trajectory images through data fusion processing.

[0044] S5. Before and after blasting, perform the observation procedures from S2 to S4 on the same borehole to be tested to obtain the three-dimensional digital core map, trajectory map and corresponding raw data of the two observations.

[0045] S6. Compare the observation data before and after the blasting to extract the location, number, opening, extension direction and distribution density of the newly added cracks;

[0046] S7. Based on the detonation pressure, charge structure and geological conditions of the blasting project, analyze the fracture development law and deduce the optimal hole radius and blasting parameter combination.

[0047] S8. Feed back the optimal combination of blasting parameters to the next cycle of blasting design.

[0048] Furthermore, in S1, the design of the miniature 3D scanning machine fully considers the needs of actual engineering environments. Its compact size allows it to easily enter various types of boreholes for observation. The surround-view camera array can capture details of the borehole wall from all directions. The laser scanning unit can accurately obtain the 3D coordinate information of the borehole wall. The inertial measurement unit ensures the machine's stable positioning in complex drilling environments. The autonomous navigation system realizes the automatic movement control of the machine. Its waterproof, dustproof, and vibration-resistant performance, as well as its ability to adapt to a wide temperature range, ensure the reliable operation of the machine under harsh working conditions.

[0049] In S2, the flexible guidance device is ingeniously designed to adapt to boreholes with different degrees of curvature, accurately guiding the 3D scanning machine to the position to be measured. The activation of the autonomous navigation system enables the machine to move at a constant speed along a preset path, providing a stable foundation for subsequent data acquisition.

[0050] In S3, the surround-view camera array, laser scanning unit, and inertial measurement unit work together. The panoramic images of the borehole wall captured by the surround-view camera array provide intuitive visual information for subsequent analysis, while the three-dimensional coordinate point cloud collected by the laser scanning unit constructs an accurate geometric model of the borehole wall. The inertial measurement unit corrects the machine's attitude and depth positioning in real time, ensuring the accuracy and consistency of the data.

[0051] In S4, the ground terminal has advanced data fusion processing technology, which can efficiently fuse received images with point cloud data to generate high-precision three-dimensional digital core images and borehole trajectory maps, providing detailed data support for subsequent fracture analysis.

[0052] In S5, by performing the same observation process on the same borehole before and after blasting, key data from the two observations were obtained, providing a basis for comparative analysis of fracture changes.

[0053] In S6, by comparing the observation data before and after the blasting, various parameters of the newly added fractures can be accurately extracted, providing a quantitative basis for the analysis of fracture development patterns.

[0054] In S7, by combining relevant parameters of blasting engineering and geological conditions, the development law of fractures is analyzed in depth, and the optimal hole radius and blasting parameter combination derived from it has high practical value.

[0055] In S8, the optimal combination of blasting parameters is fed back to the blasting design of the next cycle, realizing continuous optimization and improvement of blasting engineering.

[0056] The surround-view camera array consists of 4 to 8 CMOS sensors, which are evenly distributed in a circle, and the resolution of a single CMOS sensor is no less than 1080P.

[0057] Furthermore, 4 to 8 CMOS sensors evenly distributed in a circle can be fused using an image stitching algorithm to form a 360° panoramic image of the borehole wall without blind spots. Each sensor has a resolution of at least 1080P, capable of clearly capturing the texture features of minute cracks (≥0.1mm wide) in the borehole wall, providing high-definition visual data for subsequent crack identification. Compared to traditional unidirectional cameras, this array avoids blind spots in localized areas of the borehole wall. The redundant design of multiple sensors ensures that if one sensor temporarily fails, the remaining sensors can still cover most of the observation range, improving system reliability.

[0058] Additional information on suitable scenarios: When the borehole diameter is small (e.g., ≤50mm), prioritize using 4 CMOS sensors (to reduce device size). If the borehole diameter is large (e.g., ≥80mm), use 8 sensors (to improve the detail density of the panoramic image), thus adapting to the observation needs of boreholes of different sizes.

[0059] The laser scanning unit is a single-line or multi-line rotating lidar with a scanning frequency of not less than 20Hz and a ranging error of not more than ±2mm.

[0060] Furthermore, single-line rotating lidar is suitable for scenarios with moderate requirements for borehole wall contour accuracy (such as conventional hard rock borehole observation), and features low cost and low power consumption. Multi-line rotating lidar, on the other hand, is suitable for scenarios with high requirements for the three-dimensional morphological details of the borehole wall (such as soft rock boreholes with complex fracture development), and can simultaneously acquire multi-dimensional point cloud data, improving the accuracy of borehole wall modeling.

[0061] Performance and Function: The scanning frequency is no less than 20Hz, which can match the moving speed of a miniature 3D scanner (0.05~0.2m / s), ensuring that each section of the borehole wall is fully scanned and avoiding data loss. The ranging error does not exceed ±2mm, which can accurately capture the changes in the opening of the borehole wall cracks, providing reliable spatial coordinate data for the subsequent quantitative extraction of crack parameters.

[0062] When the borehole diameter is ≤60mm, a smaller single-line lidar should be preferred. If the borehole diameter is ≥80mm and it is necessary to observe minute cracks (opening ≤0.5mm), a multi-line lidar should be selected to balance the size of the device and the observation accuracy.

[0063] Data transmission in S4 can be either wired or wireless. Wireless transmission uses 5G or LoRa communication protocols, supporting remote control and real-time data analysis, with a data transmission latency of no more than 50ms.

[0064] Furthermore, wired transmission is suitable for drilling depths ≤50m and environments with strong electromagnetic interference, offering advantages such as stable data transmission and strong anti-interference capabilities. Wireless transmission, on the other hand, is suitable for drilling depths or scenarios requiring remote control, and the appropriate protocol can be selected based on actual needs: the 5G protocol supports high-speed, low-latency real-time remote control, while the LoRa protocol is suitable for long-distance (≤10km) and low-power data backhaul.

[0065] Protocol performance characteristics: The 5G protocol can achieve a transmission rate of over 1Gbps, meeting the real-time transmission requirements of high-definition images of borehole walls and 3D point clouds. The LoRa protocol has a long communication distance and low power consumption, making it suitable for blasting engineering scenarios in the field without network coverage, and ensuring stable data transmission.

[0066] Transmission performance assurance: Data transmission latency does not exceed 50ms, ensuring that the ground terminal responds to the control commands of the miniature 3D scanning machine in real time. At the same time, the real-time data analysis function can quickly provide feedback on the borehole wall status, avoiding observation deviations or machine malfunctions caused by latency.

[0067] The flexible guide device is made of high-strength nylon braided material, and its diameter can be adaptively adjusted within the range of 8 to 50 mm to adapt to boreholes of different diameters. It can also guide the 3D scanning machine through borehole sections with a bending angle of no more than 15°.

[0068] Furthermore, the flexible guide device consists of a high-strength nylon braided layer and an internal steel wire skeleton, providing basic flexibility and tensile strength. Its diameter can adaptively shrink / expand within the range of 8–50 mm to accommodate different boreholes, eliminating the need for separate devices for different boreholes and reducing equipment costs. The internal steel wire skeleton further enhances the device's tensile and deformation resistance. In complex drilling environments, when the device is subjected to external pulling or squeezing forces, the steel wire skeleton effectively disperses stress, preventing excessive deformation or breakage and ensuring the smooth passage of the 3D scanning machine. Simultaneously, the flexible guide device's surface undergoes special treatment, providing excellent anti-static properties. This prevents dust adsorption from the borehole walls due to static electricity in dry drilling environments, avoiding impacts on the device's normal operation and the scanning machine's detection accuracy. In addition, the device is lightweight, with a low overall weight, which does not place an excessive burden on the 3D scanning machine's propulsion system, thus improving the machine's movement efficiency and stability within the borehole. The high-strength nylon braided material combines flexibility and abrasion resistance, with a tensile strength exceeding 500 MPa. It adapts to the changing shape of curved sections in the borehole while resisting frictional wear from the borehole wall rock, ensuring the reusability of the device (wear rate ≤2% after a single operation). The flexible guide device functions similarly to a connecting rod adapted to the borehole sight.

[0069] The practical value of the curved section guide: It can traverse borehole sections with an angle of ≤15°, solving the problem that traditional rigid connecting rods cannot be adapted to directional drilling and slightly curved drilling. It is especially suitable for borehole observation in complex geological areas (such as rock strata with well-developed joints), expanding the engineering application scope of the device.

[0070] The autonomous navigation system has a built-in preset trajectory planning module and a real-time obstacle avoidance module. Through the coordinated feedback of the inertial measurement unit and the laser scanning unit, the movement speed of the 3D scanning machine can be adjusted within the range of 0.05 to 0.2 m / s, and the movement deviation does not exceed ±3 mm / m.

[0071] Furthermore, the preset trajectory planning module can generate the optimal movement path based on the pre-detection parameters of the borehole (such as borehole depth and diameter), avoiding repeated scanning or missed areas. The real-time obstacle avoidance module automatically adjusts the movement direction based on the information of obstacles (such as protruding debris) on the borehole wall fed back by the laser scanning unit, preventing the machine from jamming or being damaged. The inertial measurement unit captures the machine's attitude deviation in real time, and the laser scanning unit simultaneously collects spatial data of the borehole wall. After the two data are fused, they are transmitted to the autonomous navigation system, which dynamically adjusts the movement speed and direction, realizing closed-loop control from attitude perception, data feedback to trajectory correction.

[0072] Engineering value of performance parameters: Adjustable speed of 0.05~0.2m / s: Low speed is suitable for fine scanning of complex borehole walls, while high speed is suitable for efficient observation of conventional borehole sections, balancing accuracy and efficiency. A movement deviation of ≤±3mm / m ensures stable movement of the machine along the borehole axis, avoiding incomplete borehole wall data acquisition due to path deviation and guaranteeing the modeling accuracy of the 3D digital core image.

[0073] The actual working principles of the preset trajectory planning module and the real-time obstacle avoidance module are as follows:

[0074] When the preset trajectory planning module is working, it first receives the pre-detection parameters of the borehole, which include key information such as borehole depth and diameter. Internally, the module uses advanced algorithms to comprehensively analyze these parameters and generates an optimal movement path based on the analysis results. This path fully considers the actual conditions of the borehole, effectively avoiding repeated scanning of certain areas or omission of parts during subsequent scanning, thereby improving the comprehensiveness and efficiency of the scanning work.

[0075] The real-time obstacle avoidance module relies on a laser scanning unit. This unit continuously scans the hole wall, transmitting information about obstacles such as protruding debris to the real-time obstacle avoidance module. Upon receiving this information, the module quickly initiates an automatic adjustment program, adjusting the machine's direction of movement based on the obstacle's position and size. This ensures the machine can smoothly navigate around obstacles, preventing jamming during movement and avoiding damage from collisions, thus guaranteeing safe operation in complex hole wall environments.

[0076] The data fusion processing in S4 uses an image-point cloud registration algorithm based on Kalman filtering to fuse the texture information of the two-dimensional image with the spatial location information of the three-dimensional point cloud, generating a three-dimensional digital core image of the borehole with a resolution of no less than 0.1 mm / pixel.

[0077] Furthermore, the Kalman filter registration algorithm solves the problem of information heterogeneity between two-dimensional images (containing only texture) and three-dimensional point clouds (containing only spatial coordinates) by predicting and correcting the errors of image-point cloud data in real time, and achieves high-precision alignment between the two. This enables the texture details of the image to be mapped onto the spatial structure of the point cloud, restoring the true shape of the hole wall.

[0078] The core value of data fusion: A single two-dimensional image cannot reflect the three-dimensional undulation of the hole wall, and a single three-dimensional point cloud lacks texture details. The fused data not only retains the visual characteristics of the crack (such as width and direction) but also has spatial location information (such as depth and distribution density), providing a complete data foundation for the quantitative analysis of crack parameters.

[0079] A three-dimensional digital core image with a resolution of at least 0.1 mm / pixel can clearly show minute cracks with a width of ≥0.1 mm, meeting the accuracy requirements for crack observation after hard rock blasting (hard rock cracks are mostly in the range of 0.1 to 2 mm), and avoiding the omission of key crack information due to insufficient resolution.

[0080] In S6, a deep learning semantic segmentation model is used to compare the observation data before and after the blasting, automatically identify newly added crack areas, and the quantitatively extracted crack opening measurement accuracy is no less than ±0.05mm, and the extension direction measurement accuracy is no less than ±2°.

[0081] Furthermore, the deep learning semantic segmentation model can automatically identify newly added crack areas after blasting by pre-training the crack features (such as texture abrupt changes and grayscale differences) of the borehole wall images before and after blasting. This is different from the traditional manual comparison method, which not only reduces subjective errors but also improves the comparison efficiency by more than 40%.

[0082] A fracture opening accuracy of ±0.05 mm can capture the changes in minute fractures after hard rock blasting (such as an opening of 0.1–0.5 mm), providing precise data for assessing the energy transfer efficiency of blasting. An extension direction accuracy of ±2° can clearly define the fracture propagation trend, assist in analyzing the stress distribution of the rock mass, and avoid structural instability caused by fracture propagation in subsequent engineering projects.

[0083] Automated crack identification and quantitative extraction can quickly form a crack parameter database. Combined with parameters such as detonation pressure and charge structure, it can more scientifically deduce the optimal blasting scheme, solving the problem of traditional blasting effect evaluation relying on experience and lacking data support.

[0084] The geological conditions combined in S7 include rock mass compressive strength, rock mass integrity coefficient, rock stratum dip angle and joint development density. By establishing a multivariate regression model of fracture development parameters with detonation pressure, charge structure and geological conditions, the optimal combination of blasting parameters is deduced.

[0085] Furthermore, the logic of the impact of blasting on geological conditions:

[0086] Rock mass compressive strength: Higher strength requires a higher initiation pressure to form effective fractures. Rock mass integrity coefficient: Higher integrity makes fracture propagation more difficult, necessitating optimized charge structure (e.g., interval charging). Rock stratum dip angle and joint density determine the preferred direction of fracture propagation; therefore, the borehole radius must be adjusted accordingly to prevent fractures from extending into hazardous engineering areas.

[0087] Multiple regression models quantify the relationship between fracture development parameters (such as number and opening) and detonation pressure, charge structure, and geological conditions. This eliminates the interference of single factors and accurately calculates the optimal parameter combination for different geological scenarios. For example, in high-integrity hard rock, the model can output a combination scheme of increasing the detonation pressure by 15% and using interval charges.

[0088] Compared to traditional experience-based blasting parameter design, this model shifts parameter selection from experience-dependent to data-driven, which can improve blasting effects (such as fracture uniformity) by more than 20%, while reducing explosive consumption and achieving safe, efficient and economical blasting operations.

[0089] The 3D scanning machine is also equipped with a removable battery pack, which can last for no less than 4 hours on a single charge. It supports hot-swappable replacement and has a built-in data storage module that can store no less than 100GB of observation data locally.

[0090] Furthermore, the detachable battery pack offers a single-charge runtime of ≥4 hours, meeting the drilling length requirements of most mines and tunnel projects. The hot-swappable design allows for battery replacement without power interruption, preventing interruptions to observations due to insufficient battery life and ensuring operational continuity. The 100GB local data storage module can store high-definition images and point cloud data from approximately 50 boreholes, adapting to multi-batch continuous observation scenarios. Local storage serves as a backup mechanism for data transmission; when wireless / wired transmission is interrupted, observation data is not lost and can be subsequently exported for retransmission and analysis. This design addresses the pain points of unstable power supply and limited data transmission conditions in downhole operations, enhancing the independent operation capability and data reliability of the 3D scanning machine in complex engineering environments.

[0091] The working principle and usage process of this invention are as follows: Utilizing 3D scanning technology, a laser beam is emitted to the inner wall of the borehole, and the reflected signal is received to construct a high-precision 3D point cloud model, accurately capturing the morphology, location, and distribution characteristics of blasting cracks. In terms of usage, the operator first installs the 3D scanning machine at the borehole opening. After starting the equipment, it automatically completes self-checks and calibrations. Subsequently, the machine performs a comprehensive scan of the borehole using preset scanning parameters, requiring no manual intervention. After scanning, the data is transmitted to the local storage module in real time, and also supports wireless / wired uploads to the cloud server. Finally, the point cloud data is processed and analyzed using accompanying software to generate a crack visualization report, providing a scientific basis for subsequent engineering decisions.

[0092] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to the above embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for three-dimensional observation of boreholes using micromachines to inspect blasting cracks, characterized in that, Includes the following steps: S1. Design and manufacture a miniature 3D scanning machine. The main body length of the 3D scanning machine is less than the diameter of the borehole to be measured. It is equipped with a surround-view camera array, a laser scanning unit, an inertial measurement unit (IMU), and an autonomous navigation system. It is also waterproof, dustproof, and vibration resistant, and its working temperature is adaptable to an environment of -10℃ to 60℃. S2. The three-dimensional scanning machine is sent into the borehole to be tested by a flexible guide device, and the autonomous navigation system is started to control the three-dimensional scanning machine to move at a constant speed along the borehole axis. S3. During the movement of the three-dimensional scanning machine, the surround-view camera array continuously captures panoramic images of the hole wall, the laser scanning unit synchronously collects the three-dimensional coordinate point cloud of the hole wall surface, and the inertial measurement unit corrects the attitude and depth positioning of the three-dimensional scanning machine in real time. S4. The ground terminal receives images and point cloud data, and generates high-precision three-dimensional digital core images and borehole trajectory images through data fusion processing. S5. Before and after blasting, perform the observation procedures from S2 to S4 on the same borehole to be tested to obtain the three-dimensional digital core map, trajectory map and corresponding raw data of the two observations. S6. Compare the observation data before and after the blasting to extract the location, number, opening, extension direction and distribution density of the newly added cracks; S7. Based on the detonation pressure, charge structure and geological conditions of the blasting project, analyze the fracture development law and deduce the optimal hole radius and blasting parameter combination. S8. Feed back the optimal combination of blasting parameters to the next cycle of blasting design.

2. The method for three-dimensional observation of boreholes and inspection of blasting cracks using a micromachine as described in claim 1, characterized in that, The surround-view camera array consists of 4 to 8 CMOS sensors, which are evenly distributed in a circle, and the resolution of a single CMOS sensor is not less than 1080P.

3. The method for three-dimensional observation of boreholes and inspection of blasting cracks using a micromachine as described in claim 1, characterized in that, The laser scanning unit is a single-line or multi-line rotating lidar with a scanning frequency of not less than 20Hz and a ranging error of not more than ±2mm.

4. The method for three-dimensional observation of boreholes and inspection of blasting cracks using a micromachine as described in claim 1, characterized in that, Data transmission in S4 can be either wired or wireless. Wireless transmission uses 5G or LoRa communication protocols, supporting remote control and real-time data analysis, with a data transmission latency of no more than 50ms.

5. The method for three-dimensional observation of boreholes and inspection of blasting cracks using a micromachine as described in claim 1, characterized in that, The flexible guide device is made of high-strength nylon braided material, and its diameter can be adaptively adjusted within the range of 8 to 50 mm to adapt to boreholes of different diameters. It can also guide the three-dimensional scanning machine through borehole sections with a bending angle of no more than 15°.

6. The method for three-dimensional observation of boreholes and inspection of blasting cracks using a micromachine as described in claim 1, characterized in that, The autonomous navigation system has a built-in preset trajectory planning module and a real-time obstacle avoidance module. Through the coordinated feedback of the inertial measurement unit and the laser scanning unit, the moving speed of the three-dimensional scanning machine can be adjusted within the range of 0.05 to 0.2 m / s, and the moving deviation does not exceed ±3 mm / m.

7. The method for three-dimensional observation of boreholes and inspection of blasting cracks using a micromachine as described in claim 1, characterized in that, The data fusion processing in S4 uses an image-point cloud registration algorithm based on Kalman filtering to fuse the texture information of the two-dimensional image with the spatial location information of the three-dimensional point cloud, generating a three-dimensional digital core image of the borehole with a resolution of no less than 0.1 mm / pixel.

8. The method for three-dimensional observation of boreholes and inspection of blasting cracks using a micromachine according to claim 1, characterized in that, In S6, a deep learning semantic segmentation model is used to compare the observation data before and after the blasting, automatically identify newly added crack areas, and the quantitatively extracted crack opening measurement accuracy is no less than ±0.05mm, and the extension direction measurement accuracy is no less than ±2°.

9. The method for three-dimensional observation of boreholes and inspection of blasting cracks using a micromachine according to claim 1, characterized in that, The geological conditions combined in S7 include rock mass compressive strength, rock mass integrity coefficient, rock stratum dip angle and joint development density. By establishing a multivariate regression model of fracture development parameters with detonation pressure, charge structure and geological conditions, the optimal combination of blasting parameters is deduced.

10. The method for three-dimensional observation of boreholes and inspection of blasting cracks using a micromachine according to claim 1, characterized in that, The 3D scanning machine is also equipped with a removable battery pack, which can last for no less than 4 hours on a single charge. It supports hot-swappable replacement and has a built-in data storage module that can store no less than 100GB of observation data locally.

Images