A full-automatic laser scanning and detecting system and method for roadway chamber

Abstract

This invention provides a fully automated laser scanning detection system and method for tunnels and chambers, belonging to the field of underground space exploration technology. The system includes: a drone, a multi-sensor fusion positioning module, a 3D laser scanner, an adaptive environmental perception and obstacle avoidance module, and a ground control station. The multi-sensor fusion positioning module provides high-precision pose information in environments without GPS signals; the 3D laser scanner collects laser point cloud data of the tunnels and chambers; and the adaptive environmental perception and obstacle avoidance module senses changes in dust and illumination in real time and plans a safe flight path. This invention effectively overcomes the challenges of complex environments such as GPS rejection, uneven illumination, and dust interference within tunnels through multi-source data fusion and adaptive algorithms. It achieves fully automated, high-precision, and high-efficiency 3D topographic detection in areas inaccessible to personnel, providing a reliable technical means for monitoring surrounding rock deformation and assessing risks in goaf areas.

Description

Technical Field

[0001] This invention relates to the field of underground space exploration technology, and in particular to a fully automated laser scanning exploration system and method for tunnels and chambers. Background Technology

[0002] In underground engineering projects such as mining and tunnel construction, the topographic detection of roadways, chambers, and goaf areas is crucial, primarily used for surrounding rock stability assessment, disaster early warning, and engineering quantity calculation. Traditional manual measurement methods suffer from low efficiency, poor accuracy, and high safety risks, especially in hazardous areas inaccessible to personnel (such as high-risk areas and goaf areas), where they are almost impossible to implement. In recent years, mobile 3D laser scanning technology has been applied to roadway surrounding rock deformation monitoring due to its non-contact, high-precision, and high-efficiency characteristics. However, existing technologies still face severe challenges: First, GPS signals are completely absent in roadways, making positioning difficult for mobile platforms such as drones; second, the lighting conditions in roadways are extremely uneven, ranging from complete darkness to strong light, severely affecting the robustness of visual positioning; third, the high concentration of dust generated during mining operations severely scatters laser light, resulting in numerous voids and distortions in the laser point cloud data, while also obstructing visual sensors and causing collision accidents. Although some research has attempted to combine drones with laser scanners, in complex and harsh roadway environments, their autonomy, adaptability, and data reliability are far from reaching the level of practical engineering applications. Therefore, there is an urgent need for a fully automated, highly robust laser scanning detection system that can overcome the aforementioned bottlenecks.

[0003] Unmanned aerial vehicles (UAVs) equipped with 3D laser scanning equipment have become an important technological means for underground space exploration. However, in confined environments such as tunnels and chambers, existing technologies face many bottlenecks: First, conventional laser scanners are large and heavy, which can lead to flight imbalance and excessive load when mounted on UAVs, seriously affecting flight safety and stability; second, in environments without GPS signals, with uneven lighting, and with dust interference, the positioning accuracy of UAVs drops sharply, making autonomous obstacle avoidance and directional flight difficult; third, multi-sensor fusion technology has high hardware costs and immature real-time data processing algorithms, resulting in insufficient stability and reliability of the system under complex underground working conditions. Summary of the Invention

[0004] This invention aims to solve at least one of the technical problems existing in the prior art, and proposes a fully automatic laser scanning detection system and method for tunnel chambers, so as to solve the problems of insufficient positioning accuracy, poor environmental adaptability and unreliable data quality in the prior art in complex tunnel environments.

[0005] In a first aspect, embodiments of the present invention provide a fully automated laser scanning detection system for tunnel chambers, comprising:

[0006] The drone consists of modules A, B, and C, which are fixedly connected in sequence to form a self-balancing and stable structure.

[0007] The A module includes a power supply for counterweights and a 3D laser scanner for scanning tunnel chambers to obtain laser point cloud data.

[0008] The B module includes a rotary motor and a fixed shaft. The two ends of the fixed shaft are respectively fixedly connected to the modules A and C. The stator of the rotary motor is connected to the fixed shaft, and at least three laser sensor probes are evenly arranged on the rotor.

[0009] The C module includes a control motherboard and a battery for controlling the flight of the drone;

[0010] The ground control station communicates with the UAV via a wireless link to send detection mission instructions, receive and process the laser point cloud data, and monitor the UAV's status in real time.

[0011] Preferably, the three laser probes carried by the rotor are located at 0 degrees, 120 degrees and 240 degrees respectively; the working mode of the rotor is: first rotate 120 degrees forward, then reset, then rotate 120 degrees in reverse, then reset, and so on.

[0012] Preferably, module A includes a multi-sensor fusion positioning unit. The multi-sensor fusion positioning unit adopts a tightly coupled Kalman filter algorithm, uses the high-frequency data of the inertial measurement unit as state prediction, and uses the visual feature point matching result and the laser point cloud data registration result as measurement update, so as to suppress the performance degradation of a single sensor in dusty or low-light environments.

[0013] Preferably, module A includes an adaptive environment perception and obstacle avoidance unit, which includes:

[0014] Dust sensors are used to detect the concentration of dust in the air;

[0015] An array of light sensors is used to detect the intensity of light from different directions;

[0016] Millimeter-wave radar is used to penetrate dust and detect obstacles ahead.

[0017] The control unit is used to generate flight control commands and scanning parameter adjustment commands based on the dust concentration, light intensity, and obstacle distance.

[0018] Preferably, the ground control station is also used to perform surrounding rock deformation monitoring, specifically including:

[0019] By comparing the laser point cloud data obtained in this exploration with historical laser point cloud data using a multi-scale model, the areas of surrounding rock displacement change were identified.

[0020] Based on the identified areas of surrounding rock displacement change, the deformation of the surrounding rock is calculated and a risk warning report is generated.

[0021] Preferably, the weight of the counterweight power supply is determined by the difference between module A and module C, so that the final weights of module A and module C are equal.

[0022] Secondly, the present invention also provides a fully automated laser scanning detection method for tunnels and chambers, the method being implemented based on the aforementioned fully automated laser scanning detection system for tunnels and chambers, the method comprising:

[0023] The drone took off and flew to the designated location.

[0024] The maximum spatial distance maxJS in front of the drone is detected in real time using the laser ultrasonic sensor in module A.

[0025] The perimeter distances J1, J2, and J3 are obtained by rotating the laser sensor probe of module B, and their average value JJ and the reference center point O (JJcosθ, JJsinθ) of the current flight section are calculated.

[0026] Using the maximum detected forward distance maxJS as the forward direction, and combined with the reference center point O, the vector coordinates O(JJcosθ, JJsinθ, maxJS) of the directional navigation flight detection direction are generated, and the flight continues to obtain laser point cloud data;

[0027] Module C dynamically adjusts the UAV's flight position and direction of travel based on real-time updated vector coordinates to achieve directional navigation flight;

[0028] The ground control station processes the received laser point cloud data to obtain a three-dimensional model of the tunnel chamber.

[0029] Preferably, the process of processing the laser point cloud data includes:

[0030] Outliers were removed using the DBSCAN clustering algorithm, motion trajectory error compensation and deformation correction were performed using the ICP algorithm, Gaussian filtering was applied for smoothing, and missing data were checked and filled using the radial density statistical method.

[0031] Extract the edges of the processed laser point cloud data, draw the edge curves, and generate and display a 3D model of the tunnel chamber in real time.

[0032] Preferably, the ground control station processes the received laser point cloud data to obtain a three-dimensional model of the tunnel chamber, and then further includes:

[0033] By comparing and analyzing the generated 3D model with historical models, the collapse volume of the roof in the goaf can be automatically identified, and the risk level of further collapse can be predicted.

[0034] Preferably, when the dust sensor of module A detects that the dust concentration exceeds a preset threshold, the drone's flight speed is reduced, the millimeter-wave radar takes over the obstacle avoidance function, and the number of single-point samplings of the laser scanner is increased to compensate for the signal-to-noise ratio loss.

[0035] This invention provides a fully automated laser scanning detection system and method for tunnels and chambers. Through an ABC modular self-balancing design, it fundamentally solves the problem of drone flight instability caused by heavy sensors, achieving integrated hardware. A real-time vector coordinate navigation method based on multi-source sensor fusion is proposed, enabling intelligent obstacle avoidance and directional flight in complex tunnels without relying on GPS. A complete data processing pipeline from one-dimensional data to high-precision three-dimensional models is established. Through a combination of various algorithms, the accuracy of point cloud data and model quality are effectively improved, meeting the millimeter-level detection requirements of engineering projects. Furthermore, modules A and C are fixedly connected at both ends of a fixed shaft, and the stator of the rotary motor is connected to the fixed shaft. At least three laser sensor probes are evenly distributed on the rotor. A rotation of ±60 degrees is sufficient to ensure the integrity of acquiring laser and acoustic detection data across the entire perimeter, while a rotation of ±120 degrees doubles the amount of perimeter scanning data in the flight direction. Attached Figure Description

[0036] Figure 1 This is a schematic diagram illustrating the structural principle of a fully automated laser scanning detection system for tunnel chambers provided in an embodiment of the present invention.

[0037] Figure 2 The schematic diagram of module B provided for the embodiment of the present invention;

[0038] Figure 3 This is a schematic diagram of the core steps of the directional navigation flight detection method provided in the embodiments of the present invention;

[0039] Figure 4 A schematic diagram of the main steps of the laser point cloud data acquisition and 3D modeling display method provided in the embodiments of the present invention. Detailed Implementation

[0040] To enable those skilled in the art to better understand the technical solutions of the present invention, exemplary embodiments of the present invention are described below in conjunction with the accompanying drawings, including various details of the embodiments of the present invention to aid understanding. These should be considered merely exemplary. Therefore, those skilled in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0041] Where there is no conflict, the various embodiments of the present invention and the features thereof may be combined with each other.

[0042] As used herein, the term “and / or” includes any and all combinations of one or more related enumerated entries.

[0043] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used herein, the singular forms “a” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will also be understood that when the terms “comprising” and / or “made of” are used in this specification, the presence of the stated feature, integral, step, operation, element, and / or component is specified, but the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or groups thereof is not excluded. Terms such as “connected” or “linked” are not limited to physical or mechanical connections but can include electrical connections, whether direct or indirect.

[0044] Unless otherwise specified, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art. It will also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having the meaning consistent with their meaning in the context of the relevant art and the invention, and will not be interpreted as having an idealized or overly formal meaning unless expressly so defined herein.

[0045] In the technical solution of this invention, the collection, storage, use, processing, transmission, provision, and disclosure of user personal information all comply with relevant laws and regulations and do not violate public order and good morals. The use of user data in this technical solution follows relevant national laws and regulations (e.g., the "Information Security Technology - Personal Information Security Specification"). For example: appropriate measures are taken for personal information access control; restrictions are imposed on the display of personal information; the purpose of using personal information does not exceed the scope of direct or reasonable association; and explicit identity targeting is eliminated when using personal information to avoid precisely locating a specific individual.

[0046] The purpose of this invention is to provide a fully automated laser scanning detection system and method for tunnels and chambers, addressing the problems of insufficient positioning accuracy, poor environmental adaptability, and unreliable data quality in existing technologies for complex tunnel environments. To achieve the above objective, this invention provides a fully automated laser scanning detection system for tunnels and chambers, including an unmanned aerial vehicle (UAV) platform, a multi-sensor fusion positioning module, a 3D laser scanner, an adaptive environmental perception and obstacle avoidance module, and a ground control station. Through deep fusion of multi-source sensors and an environmental adaptive control strategy, the system can safely, stably, and efficiently complete detection tasks and acquire high-fidelity 3D point cloud data under extreme conditions such as the absence of GPS, low light, and high dust levels. (Reference) Figure 1 and Figure 2 Specifically, it includes:

[0047] (1) Adopting a quadcopter anti-collision propeller structure, such as Figure 1 The system consists of four propellers, surrounded by a protective ring, and includes brushless motors, batteries, and propellers; A represents the direction of flight for the drone's nose. The lift generated by the quadcopter ensures stable flight of the entire system.

[0048] (2) A is the laser detection control motherboard and data transmission module, which is mainly responsible for sending control commands to module B and collecting laser point cloud data, and carries a directional sensor and a backup power supply of a specified weight;

[0049] (3) B is a rotary motor and laser probe module, which is mainly responsible for the rotation control of three built-in laser probes. This module is mainly a hollow structure lightweight high-power rotary motor, which is mainly divided into two parts: an external rotor and an internal shaft stator. After receiving the command, the rotor can rotate around the stator. The rotor carries three built-in laser probes, which are evenly distributed in three directions: 0 degrees, 120 degrees and 240 degrees. The rotor rotates in the following manner: first rotates 120 degrees forward, then resets, then rotates 120 degrees backward, then resets, and repeats the cycle. The shaft stator is a hollow structure, and power lines and control signal lines can be run through the shaft stator.

[0050] (4) C is the UAV control motherboard and battery module, which is mainly responsible for the flight control of the UAV and is paired with the remote controller via infrared Bluetooth.

[0051] (5) Module A moves forward and is fixed to the built-in shaft stator of Module B. The built-in shaft stator of Module B is connected and fixed to Module C. The built-in shaft stator of Module A, Module B and Module C together form a self-balancing stable structure, ensuring the free and stable flight of the entire UAV system.

[0052] (6) Module B is the center of the entire flight system. In order to ensure the smooth flight of the UAV, the overall weight of modules A and C is equal.

[0053] (7) The lift of the four propellers and the power in the C module can drive the load of the entire system and ensure the smooth flight of the entire system.

[0054] The three modules, A, B, and C, together form a self-balancing rotating component capable of mounting a laser gimbal. The main features of the self-balancing rotating component are as follows:

[0055] (1) A consists of four sub-modules: a laser detection motherboard, an automatic data saving and wireless communication module, a laser / ultrasonic orientation sensor, and a backup power supply of a specified weight. The laser detection motherboard is responsible for communication and control of part B and collects laser point data. After a large amount of laser point data is analyzed and processed by the central processing unit of the laser detection motherboard, it forms laser point cloud data and is transmitted to the "automatic data saving and wireless communication module" for automatic saving and real-time wireless data transmission. The laser / ultrasonic orientation sensor is used to assist the overall UAV system in orientation and positioning flight and to assist in the processing of laser point cloud data. The backup power supply of a specified weight is used for emergency power supply and automatic saving of laser point cloud data. Its weight and size are determined by the difference between module A and module C to ensure that the two modules A and C are balanced and equal in weight.

[0056] (2) B consists of a lightweight, hollow rotary motor and a fixed shaft. To reduce flight load, a lightweight carbon fiber alloy motor is used. The stator of the motor is hollow and is fixed to the fixed shaft connecting modules A and C. The shaft is hollow and allows for wiring to pass through it, connecting the power supply and control signal lines between A and C. The rotor of the motor is fixed to a sensor module with three built-in laser probes. The sensor module consists of three built-in high-precision laser sensor probes evenly distributed around a 360-degree circumference. The power supply and signal lines of the sensor probes are appropriately extended and connected to the laser detection control mainboard of module C.

[0057] Furthermore, modules A and C are fixedly connected to the two ends of a fixed shaft, and the stator of the rotary motor is connected to the fixed shaft. At least three laser sensor probes are evenly distributed on the rotor, with the three probes located at 0 degrees, 120 degrees, and 240 degrees respectively. The rotor operates in a 120-degree clockwise rotation, then reset, then reverse 120 degrees, and reset again, repeating this cycle. Therefore, a rotation of ±60 degrees is sufficient to ensure the integrity of acquiring the entire perimeter's laser and acoustic detection data, and a 120-degree rotation doubles the amount of perimeter scanning data in the flight direction.

[0058] (3) C consists of a UAV control board and a matching battery, as well as a matching infrared Bluetooth remote controller (remote client). The UAV control board is responsible for the free flight control of the entire UAV system, connects to the quadcopter propeller, and is responsible for collecting information such as the UAV's attitude, speed, and position. It also calculates control commands based on the data and adjusts the flight state through the actuator. The matching battery mainly drives the motor to generate lift through battery power, which is the basic power source for flight.

[0059] Key points and protection points of this invention:

[0060] (1) A novel rotating laser probe UAV structural design, corresponding to Figure 1 and Figure 2 .

[0061] The main advantages are: a new type of drone that can be carried by redesigning the main frame structure of the drone (three modules A+B+C), and then adjusting the counterweight of modules A and C to achieve self-balancing rotation scanning of the drone flight system, ensuring the availability of the original drone system and the independence and reliability of the new functions; module B uses the most original miniaturized laser probe module and carbon fiber lightweight rotary motor, which greatly reduces the take-off load of the drone and does not interfere with the original software control flight system of the drone.

[0062] Module B is centrally located in the overall system design and drives multiple laser sensors for rotating scanning detection. Module B employs a simple position and angle design, fundamentally solving the problems of completeness and direct acquisition of laser scanning data. Other UAV-based laser scanning detection systems either need to avoid obstructions from the UAV fuselage or propellers, or require multiple adjustments to the UAV's flight attitude to avoid these obstructions and perform multiple fill-in scans. This makes it impossible to obtain laser point detection data of the tunnel / chamber perimeter in a single flight. To ensure data integrity, the rotary motor angle θ in Module B only needs ±60 degrees of rotation to guarantee the complete acquisition of laser and acoustic detection data for the entire perimeter. A ±120 degree rotation doubles the amount of perimeter scanning data in the flight direction, providing double the redundant data for later data verification and correction. Therefore, the rotary motor angle θ ranges from ±60° to ±120°.

[0063] (2) The directional navigation flight detection method, under the premise of acoustic laser traction and real-time correction algorithm, can achieve contact-resistant autonomous intelligent flight, corresponding to Figure 3 .

[0064] The main advantages are: by employing laser / ultrasonic orientation sensors and their accompanying forward-facing reflectors, combined with a B-module carrying three laser scanning probes, the XYZ vector coordinates of the directional navigation flight detection direction are obtained as O(JJcosθ,JJsinθ, maxJS), and updated in real time based on real-time detection data, thereby achieving stable directional flight and laser scanning detection for the entire UAV system. Multi-sensor fusion technology is further optimized, integrating lidar, IMU, and wheeled odometers to achieve centimeter-level positioning and real-time mapping, improving the reliability of unmanned mining truck navigation.

[0065] (3) Method for acquiring laser point cloud data and displaying 3D modeling in tunnel chambers, corresponding to Figure 4 .

[0066] The main advantages are: Real-time analysis and processing of laser point data, combined with data from multiple sensors and GPS positioning information, is performed to correct and verify the data. This results in the generation of final laser point cloud data, which is then locally saved and wirelessly transmitted for real-time display on a large screen at a terminal computer near the remote control. Further analysis and processing, along with 3D smoothing, are performed on the upper-computer client software, forming a complete hardware and software system. Because miniaturized laser probes can detect tunnel and chamber boundaries with millimeter-level accuracy, the laser point cloud data, after analysis and optimization, achieves accuracy better than millimeters. This system achieves a detection accuracy of 1mm, classifying it as a high-precision laser point cloud detection and analysis system.

[0067] Based on the same inventive concept, this invention also provides a fully automated laser scanning detection method for tunnel chambers. The method is implemented based on the aforementioned fully automated laser scanning detection system for tunnel chambers, and includes:

[0068] The drone took off and flew to the designated location.

[0069] The maximum spatial distance maxJS in front of the drone is detected in real time using the laser ultrasonic sensor in module A.

[0070] The perimeter distances J1, J2, and J3 are obtained by rotating the laser sensor probe of module B, and their average value JJ and the reference center point O (JJcosθ, JJsinθ) of the current flight section are calculated.

[0071] Using the maximum detected forward distance maxJS as the forward direction, and combined with the reference center point O, the vector coordinates O(JJcosθ, JJsinθ, maxJS) of the directional navigation flight detection direction are generated, and the flight continues to obtain laser point cloud data;

[0072] Module C dynamically adjusts the UAV's flight position and direction of travel based on real-time updated vector coordinates to achieve directional navigation flight;

[0073] The ground control station processes the received laser point cloud data to obtain a three-dimensional model of the tunnel chamber.

[0074] refer to Figure 4 The method specifically includes the following steps:

[0075] For the large amount of laser point data transmitted from module B to module A via laser scanning detection data, the one-dimensional laser point data is converted into three-dimensional data based on the UAV's GPS positioning coordinates O(x, y, z). Then, the three-dimensional spatial coordinate data is calibrated and analyzed using the high-precision positioning coordinates from the laser / ultrasonic orientation sensor. The specific operation algorithm flow is as follows:

[0076] (1) The remote control C module controls the drone to take off and hover smoothly at the designated location O in the tunnel or chamber;

[0077] (2) Module A then sends a command to Module B to control Module B to start rotating ±120 degrees. Specifically, the three laser probes carried by the rotor are located at 0 degrees, 120 degrees and 240 degrees respectively. The working mode of the rotor is: first rotate 120 degrees forward, then reset, then rotate 120 degrees backward, then reset, and so on. At the same time, the three laser probes are turned on to detect the distances of the laser points around the UAV perimeter J1={i1, i2, ....., in}, J2={j1, j2, ....., jn}, J3={k1, k2, .....,kn}, and thus obtain the circumferential distance contours of the tunnel chamber at the starting position Jj, Jj={J1, J2, J3}.

[0078] (3) Based on the mean value of Jj, the center point of JJ and the reference plane can be O(JJcosθ, JJsinθ);

[0079] (4) Based on the laser / ultrasonic orientation sensor, the forward direction maxJS is obtained, and then the XYZ vector coordinates of the orientation navigation flight detection direction are obtained as O(JJcosθ, JJsinθ, maxJS);

[0080] (5) The UAV system flies forward based on the XYZ vector coordinates O(JJcosθ, JJsinθ, maxJS) of the directional navigation flight detection direction, and at the same time controls the B module to continuously rotate positive and negative repeatedly to continuously obtain laser point data; these one-dimensional laser point data are then analyzed and processed in the A module, and three-dimensional data conversion is performed based on the UAV GPS positioning coordinates O(x, y, z) and forward direction vector coordinates O(JJcosθ, JJsinθ, maxJS) to obtain laser point cloud data;

[0081] (6) These laser point cloud data are then stored and backed up locally on the chip, and simultaneously transmitted remotely via a point-to-point wireless network; the real-time detection laser point cloud data is automatically received in the UAV remote controller and the corresponding computer program;

[0082] (7) The process of 3D modeling and self-correction of laser point cloud data in the remote computer program is as follows:

[0083] Because laser point cloud data inherently contains labeled 3D spatial coordinates, these data can be preliminarily displayed as a 3D model of the tunnel chamber detected by laser scanning by loading corresponding 3D coordinate (x, y, z) values ​​into a constructed 3D coordinate system. Due to the large volume of laser point cloud data, and the possibility of multiple scans or repeated scans obtaining multiple data points at the same location, or the presence of laser point cloud data for a specific area, clustering algorithms such as DBSCAN are used to detect outliers. Random anomalies are identified by setting a distance threshold of 5–25 mm. Iterative nearest-point algorithms such as ICP are used for data deformation compensation. The flight trajectory error of the UAV is estimated within a time window of 0.1–3 seconds, keeping the error within 10 mm. Then, the processed data is Gaussian filtered to smooth the entire laser point cloud data. Radial density statistics are used with a 2–5 cubic millimeter resolution grid to check the point cloud density, filling in missing data parts through spatial continuity and improving the density and smoothness of the laser point cloud data. Finally, the outermost part of the laser point cloud data is extracted, the edge curve is drawn, and a three-dimensional model of the tunnel chamber perimeter is formed. Then, the three-dimensional model of the tunnel chamber detected by the UAV system and the corresponding laser point cloud data can be displayed on the client in real time.

[0084] Beneficial effects:

[0085] (1) CN202511429560.3 A method and system for UAV path planning based on laser scanning. This solution relates to the field of UAV path planning technology, and in particular to a method and system for UAV path planning based on laser scanning. A laser scanner is used to scan the tunnel environment in real time to obtain raw radial profile laser point cloud data. This solution does not fundamentally solve the problems of excessive UAV load and self-balancing during laser scanning detection.

[0086] However, this invention is mainly used in the fields of geotechnical tunnel engineering and underground space utilization. It not only solves the problem of autonomous flight path of UAVs (this invention uses multi-sensor real-time detection data to display the current forward direction vector coordinates in real time, which has strong real-time performance, high accuracy, high reliability, and the implementation method is simpler, more accurate and more reliable), but also realizes the acquisition and analysis of laser scanning detection data during flight. Finally, it can also realize the autonomous directional navigation flight of UAVs and intelligent detection of tunnel chamber perimeter data.

[0087] (2) CN201610384423.7 A high-precision three-dimensional model scanning device and method for underground roadways. This solution only carries a few sensors on the frame of an existing quadcopter UAV, which has problems such as unbalanced structural weight distribution and does not truly realize the function of flight detection. Since the existing laser scanner is large and heavy, the existing UAV system cannot carry it directly, nor can it directly acquire laser point cloud data. Therefore, this invention constructs a new system from the electronic components part, fundamentally modifying the main frame structure of the existing UAV to facilitate the carrying of the laser detection module and ensure the self-balance problem and the problem of excessive load of the flight system. It truly realizes the integrated design and engineering application of the UAV flight module and the laser sensor detection module.

[0088] (3) CN200910210189.6 A high-precision positioning method for underground space based on laser scanning and sequence coding graphics;

[0089] CN201711094418.3 Intelligent detection system and method for coal and gas outburst roadway disasters;

[0090] CN202110319716.8 An underground space exploration system based on unmanned aerial vehicles;

[0091] CN202111291409.X A UAV altitude-fixing system and method for inspecting underground coal mine roadways;

[0092] CN201711094470.9 Intelligent Unmanned Aerial Vehicle Detection System and Method for Mine Fire Disaster;

[0093] CN202311584282.X A method for predicting rockburst based on 3D scanning of tunnel shrinkage deformation by unmanned aerial vehicles.

[0094] The aforementioned existing technologies merely reassemble and integrate sensors on the basic framework of existing quadcopter drones, without addressing the specific application, modification, or reinvention of core components for tunnel chamber engineering applications. Since existing drones on the market generally have low payloads and lack targeted structural design adjustments, this invention directly addresses the engineering application background of rapid laser scanning detection within tunnel chambers. Focusing on facilitating the acquisition of laser point cloud data and enabling extensive engineering applications, it recreates the main frame structure of the drone and customizes an autonomous navigation and detection method, rather than simply upgrading, modifying, or integrating existing components.

[0095] Example embodiments have been disclosed herein, and while specific terminology has been used, it is for illustrative purposes only and should be construed as such, and is not intended to be limiting. In some instances, it will be apparent to those skilled in the art that features, characteristics, and / or elements described in conjunction with particular embodiments may be used alone, or in combination with features, characteristics, and / or elements described in conjunction with other embodiments, unless otherwise expressly indicated. Therefore, those skilled in the art will understand that various changes in form and detail may be made without departing from the scope of the invention as set forth in the appended claims.

Claims

1. A fully automated laser scanning detection system for tunnel chambers, characterized in that, include: The drone consists of modules A, B, and C, which are fixedly connected in sequence to form a self-balancing and stable structure. The A module includes a power supply for counterweights and a 3D laser scanner for scanning tunnel chambers to obtain laser point cloud data. The B module includes a rotary motor and a fixed shaft. The two ends of the fixed shaft are respectively fixedly connected to the modules A and C. The stator of the rotary motor is connected to the fixed shaft, and at least three laser sensor probes are evenly arranged on the rotor. The C module includes a control motherboard and a battery for controlling the flight of the drone; The ground control station communicates with the UAV via a wireless link to send detection mission instructions, receive and process the laser point cloud data, and monitor the UAV's status in real time.

2. The system according to claim 1, characterized in that, The rotor carries three laser probes located at 0 degrees, 120 degrees and 240 degrees respectively; the rotor's working mode is: first rotate 120 degrees clockwise, then reset, then rotate 120 degrees counterclockwise, then reset, and so on.

3. The system according to claim 1, characterized in that, The A module includes a multi-sensor fusion positioning unit. The multi-sensor fusion positioning unit adopts a tightly coupled Kalman filter algorithm, uses the high-frequency data of the inertial measurement unit as state prediction, and uses the visual feature point matching result and the laser point cloud data registration result as measurement update, so as to suppress the performance degradation of a single sensor in dusty or low-light environments.

4. The system according to claim 1, characterized in that, Module A includes an adaptive environment perception and obstacle avoidance unit, which includes: Dust sensors are used to detect the concentration of dust in the air; An array of light sensors is used to detect the intensity of light from different directions; Millimeter-wave radar is used to penetrate dust and detect obstacles ahead. The control unit is used to generate flight control commands and scanning parameter adjustment commands based on the dust concentration, light intensity, and obstacle distance.

5. The system according to claim 1, characterized in that, The ground control station is also used to perform surrounding rock deformation monitoring, specifically including: By comparing the laser point cloud data obtained in this exploration with historical laser point cloud data using a multi-scale model, the areas of surrounding rock displacement change were identified. Based on the identified areas of surrounding rock displacement change, the deformation of the surrounding rock is calculated and a risk warning report is generated.

6. The system according to claim 1, characterized in that, The weight of the power supply used for counterweight is determined by the difference between module A and module C, so that the final weights of module A and module C are equal.

7. A fully automated laser scanning detection method for tunnel chambers, characterized in that, The method is implemented based on the fully automated laser scanning detection system for tunnel chambers as described in any one of claims 1 to 6, and the method includes: The drone took off and flew to the designated location. The maximum spatial distance maxJS in front of the drone is detected in real time using the laser ultrasonic sensor in module A. The perimeter distances J1, J2, and J3 are obtained by rotating the laser sensor probe of module B, and their average value JJ and the reference center point O (JJcosθ, JJsinθ) of the current flight section are calculated. Using the maximum detected forward distance maxJS as the forward direction, and combined with the reference center point O, the vector coordinates O(JJcosθ, JJsinθ, maxJS) of the directional navigation flight detection direction are generated, and the flight continues to obtain laser point cloud data; Module C dynamically adjusts the UAV's flight position and direction of travel based on real-time updated vector coordinates to achieve directional navigation flight; The ground control station processes the received laser point cloud data to obtain a three-dimensional model of the tunnel chamber.

8. The method according to claim 7, characterized in that, The process of processing the laser point cloud data includes: Outliers were removed using the DBSCAN clustering algorithm, motion trajectory error compensation and deformation correction were performed using the ICP algorithm, Gaussian filtering was applied for smoothing, and missing data were checked and filled using the radial density statistical method. Extract the edges of the processed laser point cloud data, draw the edge curves, and generate and display a 3D model of the tunnel chamber in real time.

9. The method according to claim 7, characterized in that, The ground control station processes the received laser point cloud data to obtain a three-dimensional model of the tunnel chamber, and then includes: By comparing and analyzing the generated 3D model with historical models, the collapse volume of the roof in the goaf can be automatically identified, and the risk level of further collapse can be predicted.

10. The method according to claim 7, characterized in that, When the dust sensor in module A detects that the dust concentration exceeds the preset threshold, the drone's flight speed is reduced, the millimeter-wave radar takes over the obstacle avoidance function, and the number of single-point samplings by the laser scanner is increased to compensate for the signal-to-noise ratio loss.

Images