A humanoid robot for mining
By using a non-combustible shell, quantum-safe 5G communication, and fuel cell power on a humanoid robot, combined with a multi-posture movement mechanism, the problems of protection and communication instability in the mining environment have been solved, achieving efficient and safe mining operation capabilities.
Patent Information
- Application Number
- CN202610792256.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-06-03
- Publication Date
- 2026-11-17
AI Technical Summary
Existing humanoid robots are difficult to apply effectively in mining environments, especially in harsh conditions where they cannot provide high-level protection, are difficult to locate, have unstable communication and are vulnerable to attack, and cannot meet the needs of complex underground mining environments.
The protective design, which uses a shell, insulation layer and liquid cooling heat dissipation network made of non-combustible materials, combined with a quantum-safe 5G communication system, fuel cell power module, liquid-driven swing cylinder joint and multi-posture movement mechanism, integrates environmental monitoring module and redundancy design, and has adaptive quantum-5G hybrid communication, autonomous and manual control capabilities.
It achieves highly reliable navigation, protection, communication security, and adaptability to various terrains in mining environments, enabling it to complete complex tasks such as climbing, surveying, and rescue, thereby improving the safety and efficiency of mining operations.
Smart Images

Figure CN123103759A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of special robots and service robots, specifically to a humanoid robot for mining operations, inspections, and rescues. Background Technology
[0002] With the development of artificial intelligence and robotics, the integration of robots into all aspects of production and daily life is an inevitable trend. Currently, industrial robots are widely deployed in factories, but the application of humanoid robots is mainly concentrated in simple scenarios such as sweeping, cleaning, performance, and healthcare—applications that are not essential to production and daily life. To enable humanoid robots to enter production and daily life on a large scale, they must be able to perform tasks that ordinary robots cannot, and reach places that ordinary robots cannot reach. Furthermore, if humanoid robots can address human safety needs, including personal safety, food safety, and production safety, then they will become essential products for production and daily life, and this will be a crucial path for robots to enter production and daily life.
[0003] This invention expands the application scenarios of humanoid robots to mines. Mine environments are harsh, with narrow, rugged, and steep terrain. Existing humanoid robots are mostly used in flat, structured environments. While some robots possess single protective functions, there are no ordinary robots that simultaneously possess a heat-insulating protective shell, a basic protection index, and the ability to traverse various complex terrains. Furthermore, there are no humanoid robots with high-level protection in harsh environments, combustible gas sensors, and explosion-proof and flameproof functions in explosive environments, making it difficult to meet the needs of both surface and underground mine environments.
[0004] Traditional underground robots rely on wired or Wi-Fi / 4G communication, which has the following problems: (1) Large mining equipment and frequency converters generate strong electromagnetic interference, resulting in packet loss and bit errors in control commands, leading to malfunctions; (2) There is no GPS signal in long-distance tunnels, resulting in large cumulative errors in inertial navigation and difficulty in robot positioning; (3) Wireless control commands are easily intercepted or tampered with, and once the underground network is invaded, it may lead to mine disasters. At present, there is no solution that integrates 5G and practical quantum communication technology into underground robots and solves the above problems.
[0005] Intelligent mining robots encompass mining, tunneling, transportation, ore dressing, auxiliary operations, safety control, and rescue. Mining work is characterized by harsh environments, long working hours, heavy workloads, severe electromagnetic interference, and complex geological conditions. Therefore, a robot with features such as long endurance, high load capacity, stable operation under strong interference, multiple inspection methods, and emergency rescue capabilities after accidents is needed to perform these tasks. Summary of the Invention
[0006] Humanoid robots are those that resemble humans in appearance and behavior, possessing some or all of the characteristics of humans (shape, movement, perception, interaction, and intelligence), such as a head, torso, arms, and legs. They primarily move by walking upright, and in this upright position, they have a large height-to-slenderness ratio. The advantage of height is the ability to reach objects at higher altitudes and a better field of vision; the advantage of slenderness is the ability to move freely through complex passages and spaces. In climbing and exploration, bipedal primates have an advantage over quadrupeds. The same applies to robots. In certain complex terrains, such as rugged mountain paths or underground wells, ordinary robots and robotic dogs have difficulty reaching them; only humanoid robots that mimic human movements can traverse these terrains.
[0007] I. General settings for humanoid robots in mining: 1. Regarding the high-temperature protection module, the robot's shell is made of non-combustible or flame-retardant materials, with a heat insulation layer on the shell and a heat-resistant composite coating. A liquid cooling heat dissipation network is also added. This can block and absorb most of the heat, protecting the robot's core components.
[0008] 2. The protection index is at least IPX4. Through measures such as overall full sealing, potting and encapsulation of core components, adding waterproof and dustproof pressure covers and combined sealing rings to joints and hinges, constructing a curved maze-like dustproof and waterproof structure for movement gaps, using tempered glass and waterproof adhesive sealing for sensors and cameras, and adopting IP68 aviation waterproof plugs for interfaces, charging ports and communication ports, the protection index is increased to IP66 or IP68.
[0009] 3. A quantum-secure 5G communication system integrating quantum key distribution (the 5G mentioned in this invention refers to current advanced mobile communication technology; future 6G or higher communication technologies will also be upgraded accordingly) is configured as follows: ① Dual-channel fusion architecture: 5G handles high-throughput data (video, point cloud, device status), while quantum communication handles low-rate but highly secure core control commands and quantum sensing data. Both coexist at the physical layer and are isolated at the logical layer by a security controller, ensuring that even if 5G data is intercepted, it cannot affect the robot's core actions; ② Quantum key and command binding: Each motion command is accompanied by a one-time pad quantum key, which cannot be decrypted even if copied during over-the-air transmission; the machine... The human-end only executes actions when the key verification is successful and the instruction content logically matches the absolute position (error <1cm) currently measured by the quantum sensing unit, thereby preventing replay attacks and spurious instruction injection; ③ Adaptive quantum-5G hybrid frame: In response to downhole multipath fading and transient electromagnetic bursts, the system dynamically adjusts the pulse coding of quantum signals (such as BB84 protocol variants) and the mapping of 5G resource blocks to ensure that quantum signaling can still be reliably demodulated even in an environment where electromagnetic noise is 20dB higher than that of traditional communication; ④ Fault-safety mechanism: If the quantum channel is compromised or the quantum sensing unit fails, the robot automatically enters a locked-reporting-standby state and cannot be moved by any external signal unless a human arrives on-site for physical reset.
[0010] 4. The power module is a fuel cell. ① Ultra-long battery life and rapid refueling: The energy density of fuel cells is much higher than that of lithium batteries, which can support humanoid robots to work continuously for a long time on a large scale in mines; refueling only takes 3-5 minutes, much faster than charging, significantly improving work efficiency; ② Low temperature resistance and all-weather adaptability: The temperature of the mining environment is low, and the power decay of fuel cells is small at low temperatures, without the need for additional battery heating, making it highly adaptable; ③ Environmentally friendly and pollution-free: The only emission is water, which is suitable for underground environments such as coal mines that are sensitive to exhaust gas, avoiding secondary pollution or explosion risks; ④ Load insensitive: Increasing the fuel storage has little impact on the weight of the whole machine, while the increase in battery weight will significantly affect the energy consumption of movement.
[0011] 5. The load-bearing joints utilize hydraulically driven swing cylinders. ① Extremely high power density: Hydraulic drive outputs a much greater torque than a motor in the same volume, enabling humanoid robots to carry heavy objects, break rocks, or climb steep slopes; ② Shock resistance and overload protection: Liquids have low compressibility and can absorb transient impact loads, avoiding damage to gears or motors. The hydraulic system itself can release pressure, protecting the joint structure; ③ Low-speed, high-torque stability: Precise operations in mines require low speed and high torque. Hydraulic-driven swing cylinders have no reducer backlash, resulting in smoother control and eliminating the low-speed crawling phenomenon common with motors; ④ High integration: The swing cylinder directly converts hydraulic energy into rotational motion, eliminating the need for connecting rods / reducers, reducing joint volume and potential failure points, making it suitable for the compact biomimetic structure of humanoid robots.
[0012] 6. Environmental monitoring module (dust + wind speed sensor) Real-time dust concentration monitoring can provide early warning of coal dust explosion risks (automatic alarm when the lower limit is reached); wind speed sensors monitor ventilation efficiency to prevent gas accumulation (linking the ventilation system when the wind speed is too low). Explosion-proof compatibility: The dust sensor adopts the laser / light scattering principle and has a mature intrinsically safe design; the wind speed sensor is a hot film type, with no risk of sparks, and also functions as a wind direction sensor. It has high accuracy, small size, and meets the explosion-proof standards for mining.
[0013] 7. Methods for humanoid robots to adapt to narrow and complex terrain: (1) The legs are equipped with a composite mobility mechanism, and the feet integrate foot-type, wheel-type and track-type drive structures, which can be used alone or in combination. They are combined with the limb electric drive and hydraulic drive control system to take into account obstacle crossing, high-speed movement, strong grip and muddy road passage performance.
[0014] (2) Cameras and distance sensors are installed on the outside of the legs and arms. When the field of view of the sensing device on the top of the body is blocked, the navigation can be taken over by the side sensor through visual servo switching and the trajectory can be corrected in real time.
[0015] (3) The control system has a built-in multi-posture switching program, which can freely switch between upright humanoid, four-legged bionic, four-wheel driving, and crouching low center of gravity two-wheel postures, adapting to various terrains such as mountain roads, dense forests, ravines, steps, and wading.
[0016] (4) It integrates SLAM algorithm and obstacle avoidance path planning strategy, supports remote manual control and fully automatic autonomous movement, and can realize a variety of actions such as side turning, bending, crawling, climbing and hoisting; Climbing is accomplished using a robotic arm paired with an end effector. The robotic arm can be of fixed length or telescopic, and the end effector can be equipped with dexterous hands, calipers, grippers, soft claws, etc. The telescopic arm can grab distant supports to assist in climbing. Lifting can be accomplished by docking with an external aircraft through a pre-reserved interface on the fuselage, and the aircraft will transport the equipment to the work area.
[0017] The robot body is designed with pre-installed lifting lugs for drones. The drones are equipped with specialized lifting devices such as lifting cages and ring slings, and the lifting devices have self-locking and quick-unlocking structures.
[0018] II. Redundancy Design The humanoid robot body can be redundantly designed in three dimensions: degrees of freedom, energy and power, and drive control. 1. Redundancy of degrees of freedom: The range of motion and rotation angle of each joint are greater than the normal range of motion of the human body, which makes it easy to pass through narrow gaps and get up and get out of trouble on its own after falling.
[0019] 2. Energy redundancy: In addition to the main energy source, a detachable backup energy storage power source is added to ensure continuous operation under extreme conditions.
[0020] 3. Drive redundancy control: Set a safety threshold for leg force. When the load on a single leg exceeds the limit, the multi-limb collaborative escape logic is automatically activated, relying on the traction of other limbs to smoothly escape the stuck state.
[0021] 4. The hip and elbow joints adopt a large-angle bidirectional rotation structure to further improve the flexibility of movement.
[0022] 5. The backup power source uses a quick-install and removable energy storage battery, which is suitable for replacement and replenishment in multiple scenarios.
[0023] 6. The safety threshold for leg stress is set at more than 30% higher than the standard for traction force during normal walking, thereby improving the safety of operations in complex terrain.
[0024] III. About the Software This invention adopts a cloud-based collaborative modular kernel architecture, breaking down operational functions into independent dedicated functional kernels and storing them centrally in the cloud. The device is equipped with a lightweight basic operating base locally, eliminating the need for pre-installation of full model data, i.e., a local AI + cloud mode. When executing corresponding tasks, the cloud distributes dedicated core kernels on demand and dynamically and incrementally loads supporting algorithm resources, enabling instantaneous activation of functions in multiple scenarios, effectively reducing local storage usage, and improving robot adaptability and operational efficiency.
[0025] The application software adopts a layered architecture design, which includes the following layers from bottom to top: hardware physical layer, hardware driver layer, operating system kernel layer, business logic algorithm layer, and human-computer interaction application layer. Each layer calls from top to bottom and supports each other from bottom to top. The layers are decoupled from each other and the interfaces are standardized, which facilitates function expansion, module replacement and later iteration and upgrade.
[0026] (1) Hardware physical layer The physical hardware layer provides the physical hardware carrier for the entire software system, including the main control chip, storage unit, power management unit, communication module, SLAM positioning module, ultrasonic detection module, various environmental perception sensors, motion execution motors, lettering nozzle assembly, and human-machine display peripherals mounted on the robot. It provides the hardware foundation for data acquisition, processing, instruction execution, and information interaction for the upper-layer software.
[0027] (2) Hardware driver layer The hardware driver layer is located between the hardware physical layer and the operating system kernel layer. It encapsulates the low-level drivers for various hardware, including sensor drivers, ultrasonic acquisition drivers, communication serial port drivers, motor motion drivers, nozzle control drivers, positioning and navigation peripheral drivers, etc. It realizes the conversion and adaptation of hardware electrical signals, raw data and system standard instructions, provides a unified calling interface to the upper layer, and shields the underlying differences between different hardware models.
[0028] (3) Operating system kernel layer The operating system kernel layer adopts an embedded real-time operating system or HarmonyOS / Linux embedded system, which is responsible for the whole machine task scheduling, memory management, process control, network protocol adaptation, file system management and unified allocation of hardware resources; it adapts to various hardware drivers downwards, and provides a stable operating environment, communication interface and system service support for the business logic algorithm layer upwards.
[0029] (4) Business logic algorithm layer The business logic algorithm layer is the core processing layer of the robot software. It integrates SLAM simultaneous localization and mapping algorithms, complex terrain path planning algorithms, anchor bolt loosening, support thickness, and top plate delamination waveform recognition algorithms, multi-scenario task adaptation logic, environmental data threshold judgment logic, multi-module collaborative scheduling logic, and dedicated control logic for fire fighting / inspection / search and rescue scenarios. It does not rely on underlying hardware and interface interaction, and independently completes data parsing, logical judgment, algorithm calculation, and task decision-making.
[0030] (5) Human-computer interaction application layer The human-computer interaction application layer is the topmost interaction and display layer, responsible for displaying the human-computer operation interface, configuring task parameters, visualizing the running status, displaying sensor data waveforms, and pushing work logs and alarm information. It only undertakes the functions of interface display, command reception and information presentation, and does not participate in the underlying algorithm calculation and direct hardware control, so as to realize intuitive human-computer interaction.
[0031] To achieve excellent sealing and explosion-proof performance, the human-machine interface is generally not set on the robot body but is placed on the remote control.
[0032] The specific technical solution of this invention is as follows: This invention provides a humanoid robot for mining, comprising: a core processing unit, a mobility module, an environmental perception module, a work execution module, an environmental monitoring module, a protection module, a power module, a positioning and navigation module, and a communication module.
[0033] The core processing unit is electrically and / or signal-connected to the mobility module, environmental perception module, environmental monitoring module, power module, positioning and navigation module, and communication module, respectively. The robot's shell is made of non-combustible or flame-retardant materials, and the overall protection rating is at least IPX4. The mobility module has pre-stored multi-pose switching programs and programs for navigating various complex terrains.
[0034] Furthermore, the robot satisfies at least one of the following technical features: (A) The power module is a fuel cell; (B) The load-bearing joint of the power module adopts a hydraulically driven swing cylinder or a combination of a hydraulically driven swing cylinder and a geared motor; (C) The environmental perception module includes a gravimeter and / or a magnetometer to provide the robot with absolute position and geological perception signals that are independent of the navigation system and immune to electromagnetic interference; (D) The protection module also includes a heat-absorbing layer located on the outer shell, which includes a cooling pipe network that uses fire-fighting water as cooling water; and the communication module adopts encryption measures.
[0035] Furthermore, the aforementioned mining humanoid robot also includes at least one of the following items: (1) The environmental monitoring module also includes a dust concentration sensor and / or a combustible gas sensor.
[0036] (2) The environmental perception module integrates one or more of the following sensors: visible light camera array, rangefinder array, lidar, laser rangefinder, depth camera, 4D millimeter-wave radar, infrared thermal imaging module, foot pressure sensor array, inertial measurement unit, and a visible light and infrared dual-spectrum camera with the lens pointing directly upwards.
[0037] (3) The mobile module and the operation execution module support autonomous control and manual operation. The human-based operation mode uses motion capture, electromyography signals or brain-computer interface to map information to the corresponding body parts of the robot, or completes it through helmet system, voice, and joystick. During movement, the position and video are transmitted back to the operation terminal, and a real-time operation map is built based on SLAM to realize collaborative planning. The human-assisted operation mode indicates the direction and task to the operator, and the robot acts autonomously.
[0038] (4) The environmental perception module also includes cameras and / or ranging sensors distributed on the legs and / or arms. When the field of view at the top of the body is blocked, the core processing unit takes over the navigation control from the sensors through visual servoing and corrects the body movement trajectory in real time.
[0039] (5) The complex terrain is supported by the program to switch between side-lying, crouching, upright, quadrupedal, four-wheeled, climbing, riding, hoisting, and winch modes. Among them, the climbing action is completed by the arm in conjunction with the end effector. The arm is of fixed length or telescopic structure. The end effector is any one of the following: dexterous hand, caliper, gripper, octopus soft claw, hook, bucket, or other reliable object that can be firmly grasped in front. The telescopic arm can grab a fixed support at a preset distance after it is extended. The riding refers to the robot imitating a worker riding and getting off a vehicle or device. The hoisting refers to the robot using a drone to hoist to the target location when working on the ground. The winch refers to the robot imitating an off-road vehicle using its built-in winch to climb steep terrain or get out of trouble. The winch is fixed to the robot's shoulder or body.
[0040] (6) The program also includes redundancy of degrees of freedom in relation to the complex terrain, namely, the shoulder joint can rotate at an angle greater than 360° and the hip joint can rotate at an angle greater than 225°.
[0041] (7) The protection module also includes an explosion-proof design, and the core electrical components are encapsulated in an explosion-proof manner.
[0042] (8) The gravimeter is a quantum gravimeter, and the magnetic field meter is a quantum magnetic field meter.
[0043] (9) The control system has a built-in multi-posture switching program, which can freely switch between upright humanoid, four-legged bionic, four-wheeled driving, and crouching low center of gravity two-wheel posture.
[0044] Furthermore, the tasks that the aforementioned mining humanoid robot needs to perform include one or more of the following: replacement, inspection, remote sensing, measurement and control, quality inspection, mine pressure monitoring, meter reading, temperature measurement, odor detection, fire extinguishing, relocation, and marking. - The replacement refers to robots replacing workers in underground work; - The aforementioned inspection involves the robot using a UWB bidirectional mobile positioning base station to capture signals from surrounding active positioning cards, enabling roll call, headcount, and blind spot personnel statistics, as well as recording and correcting violations; - The remote sensing is used by the robot to carry the gravimeter and / or magnetometer to detect the mining face or goaf, to prevent over-excavation or to detect water hazards in advance. - The measurement and control refers to mobile wind measurement and mobile monitoring, which measures wind speed, wind direction, and gas concentration at any cross section underground, or measures toxic and harmful gases in roadways that have not been ventilated for a long time. - The quality inspection includes one or more of the following: measuring the dimensions of the tunnel section, taking photos to record the quality of the metal mesh laying, measuring the anchor bolt tightness, shotcrete thickness, masonry thickness, and grouting quality; - The mining pressure refers to the measurement of the roof delamination, sidewall shrinkage, and floor bulge in the roadway. - The reading of the meter can be achieved by taking a picture of the display interface of the meter, equipment, or switch in the specified area, identifying and uploading the image, or by reading data using an NFC module; - The temperature measurement refers to detecting the temperature of the electrical equipment casing, cables, and junction boxes; - The odor refers to the abnormal gas emitted by a local fire or spontaneous combustion of coal seam, detected using a machine olfactory sensor or harmful gas sensor. - The fire extinguishing refers to immediately alarming, cutting off power, and extinguishing the initial fire upon discovering a flame or signs of fire; - The relocation refers to the dismantling, moving, and reassembling of the monitoring and control equipment at the mining face; - The mark is a slogan or mark printed at a designated location by inkjet or printing.
[0045] Furthermore, the described mining humanoid robot also includes one or a combination of the following items: (1) The environmental monitoring module also includes a sensor group, which includes a rangefinder, wind speed sensor, wind direction sensor, smoke sensor, carbon dioxide sensor, oxygen sensor, machine olfaction sensor, micro-pressure sensor, temperature sensor, humidity sensor, support quality detection module, and inkjet head, or multiple modules operating in conjunction; the support quality detection module is equipped with a detection module including an ultrasonic probe, which can detect one or more parameters in anchor bolt loosening, top plate delamination, and support thickness.
[0046] (2) Drive redundancy control for the mobile module and the environmental perception module: When the force load on a single leg exceeds the preset threshold, the multi-limb collaborative escape algorithm is automatically triggered, and the stuck limb is released by the traction force of the remaining limbs.
[0047] (3) The power module also includes a supercapacitor as a power buffer for the fuel cell.
[0048] (4) The fuel cell is a hydrogen fuel cell, and the hydrogen storage method is solid hydrogen storage.
[0049] (5) The sensor group is fixed to the end of a telescopic rod. The sensor group at the top of the telescopic rod is an independent sensor system centered on the MCU. Each sensor establishes communication and electrical connection with the MCU, and the MCU establishes communication with the core processing unit. The telescopic rod is divided into ordinary type and rotary type. The rotary type has a base divided into a fixed part and a rotating part. The fixed part is connected to the robot, and the rotating part is connected to the telescopic rod. It can measure various parameters of any section in the roadway at the location. The ordinary type requires the robot to move laterally and then coordinate with the extension and retraction of the telescopic rod to measure.
[0050] Furthermore, the aforementioned quality inspection and mine pressure also include using the environmental sensing module or telescopic rod to perform precise 2D modeling of the cross-sectional dimensions of the roadway corresponding to a certain measurement point underground, as well as the deformation over time.
[0051] The steps for measuring tunnel dimensions and deformation using an environmental perception module are as follows: the robot arrives at the measurement point and is stationary; point cloud data of the tunnel cross-section is collected by scanning with LiDAR; the point cloud data is filtered, noise is reduced, and correction is performed; the tunnel outline is extracted and a cross-sectional model is fitted; and the net dimensions of the tunnel are calculated.
[0052] The steps for measuring the cross-sectional dimensions and deformation of a roadway using a telescopic rod are as follows: ① Determine the measurement point using "SLAM + underground personnel positioning system" or "SLAM + underground personnel positioning + fixed marker position correction in the roadway"; ② If the base plate has high hardness and no debris, the base plate elevation will not change during each measurement and can be used as a measurement reference point; start the telescopic rod to extend upward. When the micro pressure sensor at the top touches the tunnel roof, it is triggered. The MCU receives the signal and immediately sends a signal to the robot's wireless communication module through the sensor group's wireless communication module. The robot's core processing unit immediately stops the electric push rod and receives its displacement δ1. Assuming the next displacement is δ2, then δ2-δ1 is the deformation of the roof. ③ If the floor plate has low hardness and contains debris, it cannot be used as a measurement benchmark; a ruler needs to be sprayed or pasted on any side of the roadway; during measurement, the robot adjusts the laser line projector to be horizontal and turns it on, the laser horizontal line is irradiated onto the ruler, the robot uses a visible light camera to capture and identify the reading C1 of the laser line on the ruler, and then turns on the electric push rod to measure the displacement δ1 of the electric push rod according to the method in ②. Then the relative elevation of the roadway roof is C1+δ1+C, where C is the distance from the laser line projector to the top of the telescopic rod (before the telescopic rod extends). The result of the next measurement is C2+δ2+C, then (C2+δ2+C)-(C1+δ1+C)=C2-C1+δ2-δ1 is the deformation of the roadway roof during the two measurements; ④ There are two methods to measure the deformation of other parts of the top plate of this section. If the telescopic rod is of the ordinary type, the robot is moved below the point to be measured; if it is of the rotating type, the robot is stationary, the telescopic rod is rotated and extended, the oblique length and rotation angle are measured, and then the actual vertical deformation is derived by using the relationship of trigonometric functions. ⑤ When measuring the floor heave of a roadway, the method in ③ can also be used, in which case the telescopic rod extends downwards; ⑥ The method for measuring the shrinkage of the two sides is the same as above, except that the telescopic rod extends to both sides.
[0053] Furthermore, after performing 2D modeling of the tunnel cross-section, the existing fixed surrounding rock deformation monitoring system is used for verification; and / or, 3D modeling is performed on the basis of the 2D modeling to monitor the dynamic deformation of the surrounding rock. The method is as follows: N measuring points are arranged in the tunnel, and 2D accurate modeling is performed on each measuring point one by one to construct a 3D model of the entire tunnel at time t0. Similarly, 3D modeling diagrams at times t1…tm are constructed and merged to form a dynamic 3D diagram of tunnel deformation with time as the axis.
[0054] Furthermore, the encryption measures employed by the communication module are selected from at least one of the following sequences: low-strength encryption, medium-strength encryption, high-strength encryption, PQC hybrid mode, and pure PQC; and when an unauthorized access attempt to the communication channel is detected, the security controller can disable the motion control interface and only allow the transmission of a preset safety stop signal.
[0055] Furthermore, the aforementioned mining humanoid robot also includes one or more of the following items: (1) The inspection range of each robot is determined by SLAM and the underground personnel positioning system, or by setting up signs at both ends of the roadway and at the bends of the branch roadway to indicate the activity range. The signs are signs, spray paint, or stickers, including one or a combination of line segments, characters, QR codes, barcodes, and radio frequency modules.
[0056] (2) The combustible gas includes one or a combination of carbon monoxide, methane, gas, acetylene, natural gas, biogas, hydrogen sulfide, olefins, hydrogen, alcohol vapor, and ammonia.
[0057] (3) The leg force safety threshold is 30% or more higher than the standard for normal walking leg lifting traction force.
[0058] (4) The robot body is equipped with a lifting lug for drone lifting, or is equipped with a lifting cage or ring sling.
[0059] Furthermore, the aforementioned mining humanoid robot also includes at least one of the following: (1) The mobile module also includes a combination of foot-type, wheel-type, and track-type drive units for switching.
[0060] (2) The sensor data in the environmental perception module is provided to the core processing unit through the wireless communication module.
[0061] (3) The operation execution module also includes a fire hose interface, a fire water tank and a high-pressure fine water mist extinguishing device.
[0062] (4) The robot is also equipped with a fire extinguisher. The fire extinguisher nozzle is held by a universal clamping mechanism. The universal clamping mechanism is a multi-degree-of-freedom arm or a special clamping mechanism with no less than 2 degrees of freedom. The special clamping mechanism includes a quick-release buckle for clamping the nozzle and X-axis and Y-axis universal clamping motors.
[0063] (5) The tools used by the robot are mounted in the following ways: under normal conditions, one tool is connected to the robot's forearm and the other tools are placed in the toolbox; or a multi-tool turret is configured, with multiple tools radially connected to the turret, which is rotated to the target tool position when in use.
[0064] (6) The robot’s special design to adapt to harsh working environments includes: the charging port has a multi-seal structure; the soles of the feet are equipped with anti-slip textures; the head, torso and joints are equipped with impact-resistant buffer armor; and the robot has a built-in high-intensity light group.
[0065] (7) The protective module further includes coating the outer shell with a heat-resistant coating and / or laying a heat insulation layer; (8) The front end of the air filter of the internal combustion engine or fuel cell is equipped with a cooling pipe network to cool the hot air drawn in. The cooling water is fire-fighting water or circulating cooling water.
[0066] The beneficial effects of this invention compared to the prior art are as follows: Compared with the prior art, the mining humanoid robot provided by the present invention has the following beneficial effects: 1. Adopting a humanoid or multi-posture switching design, combined with various complex terrains through programming, it can move flexibly in narrow, rugged, and obstacle-filled underground environments, and perform a variety of actions such as side-stepping, bending, climbing, and wading, making it highly adaptable.
[0067] 2. Integrating gravimeters and / or magnetometers (especially the quantum version) can provide absolute position and geological sensing signals in downhole environments with no GNSS signals and strong electromagnetic interference, enabling highly reliable navigation and advanced geological exploration.
[0068] 3. Equipped with fire protection functions (cooling pipe network, fire water, fire extinguishing equipment, etc.) and explosion-proof design, it can extinguish initial fires and improve mine safety.
[0069] 4. Equipped with multiple environmental monitoring sensors and a retractable measuring mechanism, it can automatically complete the accurate measurement and modeling of parameters such as wind speed, gas concentration, and tunnel deformation, replacing manual labor in dangerous and arduous inspection and measurement work.
[0070] 5. Supports autonomous and manual multi-level control, and has encrypted communication and secure shutdown mechanisms to ensure control security.
[0071] 6. The power system uses fuel cells (such as hydrogen fuel cells with solid hydrogen storage) in conjunction with supercapacitors, which provides long range, fast refueling and no pollution. Attached Figure Description
[0072] Figure 1 This is a 3D view of the mine inspection humanoid robot of Example 1; Figure 2 This is the electrical principle block diagram of the mine inspection humanoid robot in Example 1; Figure 3 This is a block diagram of the principle of the quantum-secure 5G communication system integrating a humanoid robot for mine inspection with quantum key distribution, as described in Example 1. Figure 4 This is a schematic diagram of the mine inspection humanoid robot measuring wind in a tunnel, as shown in Example 1. Figure 5 This is a schematic diagram of Example 1, showing the measurement of 3D point cloud in a mine roadway using a lidar by a humanoid robot for mine inspection. Figure 6 This is a schematic diagram of the toolbox for the mine inspection humanoid robot in Example 1; Figure 7 This is a schematic diagram of a humanoid mining robot in a fully mechanized longwall face during top coal caving, as shown in Example 2. Figure 8 This is a schematic diagram of Example 2, showing a humanoid robot in a mining operation working as a roadheader operator; Figure 9 This is a schematic diagram of the humanoid robot in the mining face installing anchor bolts in Example 2; Figure 10 This is a schematic diagram of an operator remotely controlling a humanoid mining robot to drive anchor bolts from the ground, as shown in Example 2. Figure 11 Example 3 shows a mine inspection humanoid robot conducting ground surveys while being hoisted by a drone; Figure 12 This is a three-dimensional view of the giant humanoid robot for mine inspection in Example 3, in its upright position; Figure 13 This is a 3D view of the giant humanoid robot for mine inspection in Example 3, in a quadrupedal position; Figure 14 This is a 3D view of the giant humanoid robot for mine inspection in four-wheel configuration, as shown in Example 3. Figure 15 This is a 3D image of the giant humanoid robot for mine inspection using trees for assistance during its ascent, as shown in Example 3. Figure 16 This is a structural diagram of the outer shell of the humanoid robot for mine fire rescue, Example 4; Figure 17 This is a schematic diagram of the backup battery and main battery of the mine fire rescue humanoid robot in Example 4, as well as the rotational degrees of freedom of the arm and the thigh.
[0073] Figure 18 This is a three-dimensional view of the mine fire rescue humanoid robot in Example 4; Figure 19 This is the electrical principle block diagram of the mine fire-fighting humanoid robot in Example 4.
[0074] exist Figure 1 In the diagram, 608 is a dry powder storage container, 609 is a driving gas cylinder, 610 is a starting gas cylinder, 652 is a solenoid valve, 613 is a nozzle, 614 is a nozzle, 615 is a box, 618 is the right arm, 643 is a dual-spectrum camera on the top of the head, and 669 is a laser line projector.
[0075] exist Figure 2 In the diagram, 1 represents the operator, 3 the microphone, 38 the joystick, 39 the integrated helmet, 51 the quantum gravimeter, 52 the quantum magnetometer, 53 the Real-Time Kinematic (RTK) system, 85 the solid-state hydrogen storage container, 656 the fuel cell + supercapacitor, 93 the hydraulic pump, 655 the communication module (5G + quantum communication), 506 the quadruped drive, 621 the camera array, 622 the infrared thermal imaging, 623 the depth camera, 624 the miniature rangefinder array, 88 the inertial measurement unit, 625 the lidar, 626 the 4D millimeter-wave radar, and 630 the wind speed sensor. 631 is a wind direction sensor, 632 is a gas sensor, 650 is the robot's core processing unit, 652 is a solenoid valve, 654 is a turntable motor, 651 is a UWB bidirectional mobile positioning base station, 653 is an electric push rod, 663 is an automatic opening and closing container (sensor group), 669 is a laser line projector, 672 is a printer nozzle, 680 is a foot pressure sensor array, 681 is a near-field communication module, 673 is an oxygen sensor, 659 is a temperature sensor, 665 is an ultrasonic rangefinder, 671 is a support quality detection module, 661 is an automatic opening and closing motor, and 659 is a wireless charging battery.
[0076] Inside the automatic opening and closing container 663: 667 is a micro-pressure sensor, 634 is a smoke sensor, 636 is a carbon monoxide sensor, 637 is a carbon dioxide sensor, 673 is an oxygen sensor, 639 is a temperature sensor, 665 is an ultrasonic rangefinder, 666 is an MCU, 658 is a wireless communication module, 633 is a dust concentration sensor, 632 is a gas sensor, 631 is a wind direction sensor, 630 is a wind speed sensor, 660 is a laser rangefinder, 659 is a wireless charging battery, 668 is an ethylene sensor, 670 is an acetylene sensor, 390 is a hydrogen sulfide sensor, 671 is a support quality detection module, 661 is an automatic opening and closing motor, and 672 is a printing nozzle.
[0077] The four modules inside the automatic opening and closing container 663—micro-pressure sensor 667, ultrasonic rangefinder 665, support quality detection module 671, and printer nozzle 672—are fixed using a turret structure. The turret's central axis is connected to the end of the telescopic rod, and the turret radially connects to these four modules. During use, rotating the turret rotates the corresponding module to its highest position. The turret structure is the same as that of turret 24 in Example 3.
[0078] exist Figure 3In the diagram, 1 represents the operator, 101 the ground control center, 102 the underground 5G base station, 103 the robot body, 201 the 5G core network (AAP / UPF), 202 the 5G server (quantum key distribution / key management), 203 the 5G baseband processing + RF front-end, 204 the quantum microwave transducer, 205 the 5G baseband processing, 206 the quantum microwave antenna transducer, 207 the quantum gravimeter / quantum magnetometer, 208 the local quantum key transmission, 506 the four-limb drive, and 509 the robot core processing unit.
[0079] exist Figure 4 In the diagram, 653 is an electric actuator, 604 is a tunnel, and 663 is an automatic opening and closing container.
[0080] exist Figure 5 In the diagram, 88 is an inertial measurement unit, 625 is a lidar unit, and 381, 382, and 383 are white strip markers painted on the side of the tunnel.
[0081] exist Figure 6 In the text, 388 stands for toolbox.
[0082] exist Figure 11 In the diagram, 61 represents a drone, 62 represents a sling, 63 represents a robot, and 64 represents a lifting lug.
[0083] exist Figure 12 In the diagram, 1 represents the operator, 2 the chair, 3 the camera, 4 the driver's cab, 5 the infrared radiation-proof double-glazed glass, 6 the infrared thermal imaging glasses, 7 the lidar, 8 the humanoid robot's neck pitch axis, 10 the torso, 11 the left shoulder joint pitch axis, 12 the right shoulder joint pitch axis, 13 the left shoulder joint yaw axis, 14 the right shoulder joint yaw axis, 15 the right arm extension cylinder, 16 the left upper arm, 17 the right upper arm, 18 the left upper arm roll axis, 19 the right upper arm roll axis, 20 the left elbow joint pitch axis, and 21 the right elbow joint pitch axis. Axis, 22 is left wrist, 23 is right wrist, 24 is turret, 26 and 27 are forearms, 29 and 30 are wheels, 31 and 32 are hydraulic calipers, 33 and 34 are spikes, 36 is bellows, 37 is microphone, 38 is joystick, 39 is integrated helmet system, 51 is left hip joint pitch axis, 52 is right hip joint pitch axis, 53 is left hip joint roll axis, 54 is right hip joint roll axis, 61 is left knee joint pitch axis, 62 is right knee joint pitch axis, 71 is left ankle pitch axis, 72 is right ankle pitch axis, 73 is electric winch.
[0084] exist Figure 13 In the middle, 33 is the left spike and 34 is the right spike.
[0085] exist Figure 14 In the diagram, 29 is the left wheel and 30 is the right wheel.
[0086] exist Figure 15 In the middle, 58 is the toolbox, 15 is the right arm extension cylinder, and 32 is the hydraulic caliper.
[0087] exist Figure 16 In the diagram, 55 represents carbon fiber, 56 represents aerosol, and 57 represents cooling water pipe network.
[0088] exist Figure 17 In this configuration, 657 is the main battery and 658 is the backup battery.
[0089] exist Figure 18 In the diagram, 361 is the lifting lug, 371 is the hydraulic shears, 372 is the angle grinder, 373 is the fire extinguisher handle, 374 is the presser, 375 is the fire extinguisher, 376 is the water pipe or fire extinguisher nozzle, 377 is the quick-release buckle, 378 is the nozzle, and 379 is the nozzle universal clamping mechanism.
[0090] exist Figure 19 In the diagram, 9 is the remote control, 25 is the integrated helmet system, 84 is the satellite navigation and positioning system, 371 is the hydraulic shears, 372 is the angle grinder, 374 is the presser, 385 is the X-axis universal clamping motor, 386 is the Y-axis universal clamping motor, 655 is the wireless communication module, 621 is the camera array, 622 is the infrared thermal imaging, 623 is the depth camera, 624 is the ultrasonic rangefinder array, 625 is the inertial measurement unit, 638 is the lidar, 626 is the 4D millimeter-wave radar, and 632 is... Gas sensor, 636 is carbon monoxide sensor, 637 is carbon dioxide sensor, 390 is hydrogen sulfide sensor, 382 is life detector, 380 is the core processing unit of general fire-fighting humanoid robot, 506 is limb drive, 680 is foot pressure sensor array, 657 is main battery, 658 is backup battery, 236 is booster pump, 630 is wind speed sensor, 631 is wind direction sensor, 639 is temperature sensor, 640 is sound and light signal generator, and 641 is walkie-talkie. Detailed Implementation
[0091] 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. In the description of the present invention, the terms "upper," "lower," "front," "rear," "left," "right," "top," "bottom," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the present invention. It should also be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," and "link" should be interpreted broadly. For example, it can be a fixed connection, a detachable connection, or an integrated connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can also be a connection within two components. For those skilled in the art, these terms can be easily understood according to the specific circumstances.
[0092] Example 1 This invention provides a humanoid robot for mine inspection. The passageways in underground roadways are often narrow, such as pedestrian walkways within hydraulic supports at longwall faces, sidewalks along roadways, and pedestrian bridges over conveyor belts, with widths of only about 0.8 meters. A bipedal or wheel-legged hybrid walking mode can traverse underground steps, sleepers, and piles of rubble. Multi-jointed limbs can assist in walking close to roadway walls or climbing simple ladders, making it suitable for terrains other than flat ground where wheeled robots cannot pass. Small underground equipment, such as various sensors, probes, pipelines, valves, and inspection points, are designed according to human limb movement ranges and grasping habits, and their operations are mostly narrow-area, precise operations. Underground monitoring and control equipment and toolboxes need to be frequently moved as the working face advances; this disassembly, relocation, and reassembly work is suitable for humanoid robots. Currently, common monitoring methods mainly include fixed monitoring devices and manual inspection. Fixed monitoring devices have limited coverage because the sensors are fixed in position or follow fixed routes; robots that walk on the floor cannot reach higher areas, and robots that inspect the roof cannot reach lower areas. For complex spatial structures, such as narrow underground roadways and low pedestrian walkways reserved between machinery, especially when the ground is covered with mechanical parts, puddles, and rocks, achieving full-coverage monitoring is often difficult. The system's scalability and adaptability are poor once equipment layout or environmental conditions change. Manual inspection offers some flexibility, but poses high safety risks in certain locations, or is inaccessible to humans in other areas. Manual inspection lacks temporal and spatial continuity. In such cases, using humanoid robots can improve the continuity and coverage of inspections. Major coal mine accidents are generally related to ventilation systems. Localized gas accumulation and excessive coal dust concentration can lead to gas and coal dust explosions. Furthermore, as the working face advances, the ventilation system is constantly changing, but existing monitoring systems cannot dynamically and in real-time grasp the situation of toxic and harmful gases and airflow at any location. Although some software can simulate these conditions, the actual on-site conditions are far more complex than the ideal environment simulated by such software. Humanoid robots can walk along mine tunnels, measuring gas concentration, wind speed, wind direction, and support quality at any point on a cross-section. They can create 3D airflow maps, 3D contour maps of toxic and harmful gases, and 3D dynamic maps of tunnel deformation, and predict changes over the next 24 hours using deep learning. my country has diverse and complex mineral resources, with widespread over-exploitation and illegal mining. Using robots for inspection, due to their ability to work 24 / 7, can provide comprehensive and uninterrupted coverage of mining conditions in all mines. The advantages of humanoid robot mine inspections lie in their immediacy, continuity, and comprehensiveness. They eliminate the need for personnel to continuously measure airflow, the need to go to a wind measurement station, and the need to carry inefficient steel measuring tapes. The measured airflow is dynamic, real 3D data that does not require simulation software.For a given cross-section, the measured data represents the actual wind speed at numerous measuring points, not the average wind speed. The data is comprehensive and multi-parameter, including wind speed, wind direction, temperature, and the concentration of various gases, rather than just wind speed alone. It can measure dynamic local gas accumulation, dynamic local air leakage, coal dust concentration, and dynamic data on spontaneous combustion of coal seams.
[0093] In this embodiment of the invention, addressing the stringent explosion-proof requirements for underground mining, the robot, being a high-powered device, cannot be intrinsically safe and must be explosion-proof. Explosion-proof devices suffer from drawbacks such as large size and weight. Therefore, the robot employs the following measures: ① increasing power; ② extending battery life; ③ using high-torque hydraulically driven swing cylinders in load-bearing joints. Hydraulically driven swing cylinders, being non-electrically driven, inherently possess explosion-proof characteristics. Other joints requiring high flexibility but with low loads utilize geared motors. For increasing power and solving battery life issues, using a fuel cell with solid-state hydrogen storage is the optimal solution, meeting the requirements of long endurance, heavy load, low temperature operation, and convenient ventilation, thus matching the mining application scenario. Lithium batteries or fuel cells with high-pressure hydrogen storage cylinders can also be used in certain special circumstances. To address the slow response of fuel cells, a supercapacitor is added as a power buffer to avoid the risk of spontaneous combustion when using lithium batteries.
[0094] To quickly deploy mining robots, remote control is essential. High-speed 5G communication is necessary. However, the operation of various tasks in the mine is crucial to all aspects of safe production, and even the slightest carelessness could lead to accidents. Remote control via 5G alone carries the risk of interference from stray radio waves and malicious control by hackers. Manipulation or interference could lead to catastrophic consequences, so encrypted communication is crucial. Employing quantum encryption technology to mitigate these risks, coupling 5G signals with quantum signals for better transmission, is highly recommended. A basic form of mining robot can replace workers in underground operations.
[0095] The reasons for needing inspection robots on site are as follows: ① The underground "one ventilation and three preventions (ventilation, gas prevention, coal dust prevention, and fire prevention)" monitoring system basically uses fixed sensors for monitoring. However, the working face underground is constantly changing, and using sensors with fixed positions to monitor constantly changing objects has the problem of incomplete monitoring, leading to poor local ventilation and the accumulation of gas or coal dust. The same applies to fire prevention; for example, temperature sensors, smoke sensors, and automatic sprinkler systems on conveyor belts can only be installed in fixed positions, but the entire conveyor belt can catch fire at any point. Existing inspection robot equipment either needs to be installed on a special fixed track on the top of the roadway, or it needs to be wheeled or tracked. Installing fixed tracks requires a large amount of manpower and cannot cover every corner, but these neglected corners are often where gas and coal dust tend to accumulate. While wheeled or tracked inspection robots are highly mobile, they have the following shortcomings: First, they are large and heavy, often blocking transport vehicles and becoming a source of traffic congestion underground. Second, while they are unmanned and powerful, the technology is not mature, posing a threat to underground pedestrians. Third, they cannot pass through complex and narrow sections because underground roadways are very narrow, such as pedestrian passages reserved in hydraulic supports of longwall faces, sidewalks on the sides of roadways, and pedestrian bridges for belt conveyors, which are only 0.8 meters wide. Fourth, they are too low, as gases such as gas and carbon monoxide usually accumulate on the roof of roadways, which tracked inspection robots cannot reach. ② The expensive inspection robots that can work continuously for 24 hours cannot be classified according to human jobs. They should be used for more tasks, such as construction quality inspection, mine pressure monitoring, and the relocation of fixed monitoring equipment and toolboxes as the working face moves. ③ Tasks that human security inspectors cannot perform should also be done by mine inspection robots. For example, during the rainy season, large amounts of water accumulate in goaf areas, especially those located above the working face. Although geophysical exploration technology has made significant progress, there is still no flexible and mobile humanoid robot capable of performing this task promptly and conveniently. Furthermore, in recent years, various coal mines, especially private mines, have experienced serious problems with mining beyond permitted layers and boundaries, even leading to the emergence of "hidden areas." Security personnel are overwhelmed, and often, as soon as inspectors arrive, the mine seals off the entrances and exits to these "hidden areas." Quantum technology can solve this problem. If a quantum gravimeter or quantum magnetometer is mounted on a robot, it can sensitively detect underground conditions. For example, if a goaf is filled with water, it will inevitably cause slight changes in gravitational acceleration g and the Earth's magnetic field. If measured on the surface, gravitational acceleration g will increase; underground, below the height of the goaf, g will decrease. Water filling the goaf will increase the magnetic field both on the ground and underground; similarly, the goaf caused by cross-layer and cross-boundary mining will inevitably cause g to decrease, and a large number of mining machines in the "dark side" will inevitably cause an abnormal increase in the Earth's magnetic field; of course, non-quantum gravimeters or magnetometers can also be used, although the accuracy is slightly worse, the cost is very low.
[0096] ④ By deploying robots equipped with gravimeters and magnetometers on the ground within the mining area and measuring at fixed points, gravity gradient maps and magnetic field intensity gradient maps of the mining surface can be obtained. Anomalies in gravity or magnetic fields can be identified on these gradient maps, often caused by goafs or mining machinery. Then, by comparing the underground and surface maps with the mining depth (thickness of the overlying strata) at each point in the mining level, and using artificial intelligence to compare with the robot's actual measurements underground, the gravitational acceleration and magnetic field intensity at any point in the mining level can be calculated. In complex terrain, robots can be quickly deployed to various measurement points using drones. Upgraded versions of gravimeters and magnetometers—quantum gravimeters and quantum magnetometers—offer even higher accuracy. Similarly, in open-pit mines, the greatest safety threats are the sudden collapse of underground goafs and the sudden instability of slopes.
[0097] A. Goaf issue When open-pit mining encounters abandoned underground goaf areas, the collapse of the goaf roof can directly trigger large-scale surface subsidence, causing personnel and equipment to fall into deep underground pits instantly. Goaf detection has always been a challenge in the industry: traditional drilling methods, involving blind drilling, rarely accurately target goaf areas; conventional geophysical methods (such as direct current methods and shallow seismic methods) are limited by near-surface interference, insufficient detection depth, and low resolution, often failing to accurately delineate the three-dimensional boundaries and filling status of goaf areas. Robots equipped with quantum gravimeters and quantum magnetometers will become active, hidden hazard detectors patrolling the mine, performing all-weather gravitational and magnetic field anomaly scans of the lower part of open-pit mine slopes, the stope floor, and adjacent areas as they move along predetermined inspection routes.
[0098] Quantum gravimeters, by detecting underground density anomalies, are essentially like equipping open-pit mines with a mobile geophysical CT scanner. Goafs, filled with air, have a significantly lower density than the surrounding rock, causing localized negative gravity anomalies; water-rich areas, with water density higher than the surrounding rock, create positive gravity anomalies. By capturing these microgravity variations, quantum gravimeters can precisely locate the boundaries and depth of goafs, distinguishing between dry and waterlogged goafs, providing crucial information for safety assessments. Their detection range extends to hundreds or even thousands of meters underground, far exceeding traditional electromagnetic methods, filling a traditional geophysical exploration blind spot in open-pit mines—the area below the surface and above the depths.
[0099] Quantum magnetometers function from another dimension: detecting magnetic anomalies. In the rock strata above mined-out areas, the content and distribution of ferromagnetic minerals (such as magnetite) are often altered by mining activities, generating detectable magnetic anomaly signals. More importantly, open-pit mines may contain buried abandoned mining equipment, discarded metal components, and even unexploded explosives; these ferromagnetic materials will produce significant local magnetic anomalies in the background magnetic field. Quantum magnetometers can accurately separate weak target magnetic signals in complex, highly interfering backgrounds and, using vector measurement capabilities, calculate the target's burial depth, orientation, and approximate mass, providing precise coordinates for hazard identification and mitigation.
[0100] The synergistic effect of the two instruments enables precise three-dimensional characterization of the goaf: the gravimeter determines the density anomaly and its size, while the magnetometer locates magnetic markers and metal obstacles. The two instruments complement each other, forming a panoramic view from density anomalies to magnetic characteristics, making hidden disaster-causing bodies nowhere to hide.
[0101] B. Slope stability problem As the depth and scope of open-pit coal mining continue to expand, geological conditions become increasingly complex, and slope stability issues become more prominent. Traditional manual inspection methods are no longer sufficient to meet the needs of safe production.
[0102] Another crucial role of quantum gravimeters in open-pit mine safety inspections is their ability to detect early signs of rock mechanical imbalances caused by mining operations. As open-pit mining deepens, the stress distribution of the surrounding rock on the slope is dynamically changing. Micro-fractures and opening of structural planes in stress concentration zones cause changes in local density and porosity, leading to minute fluctuations in the gravitational field. Quantum gravimeters, with an accuracy of eight decimal places, can transform these minute changes into quantifiable disaster warning indicators. Combined with AI modeling, this can predict risks such as collapses. Early warning signals of slope instability may appear only hours or even minutes before an accident, and quantum gravimeters can achieve near real-time anomaly detection.
[0103] Quantum magnetometers also play a unique role in this scenario. Under stress, the magnetic domain structure and alignment of magnetic minerals (such as magnetite and pyrrhotite) in the slope rock mass change, resulting in measurable magnetic anomalies. This stress-induced magnetization effect provides independent physical precursor information for slope instability. When a quantum gravimeter detects an anomaly in the gravity field, data from quantum magnetometers in the same area are simultaneously retrieved and cross-referenced. This effectively eliminates environmental interference such as rainfall and temperature changes, improving the reliability and accuracy of the early warning.
[0104] Traditional slope monitoring methods (slope radar, GNSS, crack gauges, etc.) primarily measure surface displacement, lacking effective direct monitoring methods for deep underground deformation and stress states. Quantum gravimeters and quantum magnetometers precisely fill this critical gap in deep monitoring, constructing a three-dimensional monitoring network from the surface to the depths. By mounting quantum gravimeters and quantum magnetometers on robotic platforms and integrating them into existing space-air-ground integrated monitoring systems, blind spots in manned monitoring and coverage areas of fixed equipment can be filled, enabling proactive, real-time, and multi-dimensional monitoring of open-pit mine slopes, thus moving mine safety management towards proactive prevention.
[0105] Robot platform empowerment – the technological advantages of “active patrolling” + “multimodal collaboration” Quantum gravimeters and quantum magnetometers are merely sensors. Only when mounted on a robotic platform can static detection be upgraded to dynamic, full-domain inspection, allowing the two technologies to work synergistically and achieve a combined benefit greater than the sum of its parts.
[0106] Firstly, it achieves autonomous, full-area coverage, filling blind spots for manual inspections. Equipped with dual quantum sensors, the robot can autonomously inspect along a planned route without human intervention, covering areas that are traditionally difficult for manual personnel to reach or are too dangerous to be occupied in, such as steep slopes, high-risk subsidence areas, goaf subsidence areas, and deep mining areas. The robot's walking or wheeled movement detection method enables high-density, grid-based data collection, improving the completeness and continuity of the detection data compared to manual spot checks and experience-based judgments.
[0107] Secondly, the combination of static and dynamic approaches significantly improves detection efficiency. Mounting sensors on mobile robots allows for the capture of gravity and magnetic data, enabling higher-density measurements and dramatically improving efficiency compared to traditional static station-based detection methods. Simultaneously, the robot can integrate other auxiliary sensors (LiDAR, visible light cameras, infrared thermal imaging, etc.) to construct a multi-dimensional physical field collaborative detection system with quantum gravimeters and quantum magnetometers, achieving comprehensive visibility from topography to underground structures. Remote sensing satellites and UAV aerial surveys provide broad-area background information, while the robot undertakes detailed near-surface detection and high-risk area inspection tasks, realizing a seamless, multi-level, three-dimensional monitoring network integrating air, space, and ground.
[0108] Third, human-machine collaboration significantly enhances inherent safety. In the past, the detection of such hidden disaster-causing bodies relied on scattered deployment and human experience-based judgment, resulting in limited coverage and personal safety risks. Robots equipped with dual quantum sensors conduct all-weather, all-time, and all-terrain safety inspections, freeing personnel from high-risk detection operations and fundamentally preventing accidents involving detection personnel in underground hazard areas, high slope danger zones, and other similar scenarios.
[0109] ⑤ By importing the aforementioned gravity gradient map or magnetic field gradient map into the robot, the robot can perform actions that can only be performed at fixed points. For example, for a blasting robot, the detonation point must be more than 1 meter away from the working face to detonate. If the operator orders the robot to detonate at a position of 50 meters, the robot will deduce its actual coordinates through the coordinates of the working face and the gravity gradient map or magnetic field gradient map of its location. Finding itself only 50 meters from the blasting face, it will refuse to perform the task. ⑥ By increasing the robot's protection rating to IP68, the robot can enter flooded tunnels to carry out rescue operations. After discovering trapped personnel, the coordinates of the trapped personnel should be promptly reported to the rescue command center, and food, water, and oxygen should be supplied to the trapped personnel.
[0110] In this embodiment of the invention, a heat insulation layer is laid on the robot's shell to ensure high-temperature resistance during underground firefighting. It is also equipped with combustible gas sensors required for underground inspections, and the protection rating has been upgraded to IP68. The robot features an explosion-proof design, with core electrical components using explosion-proof encapsulation and obtaining ExdIIBT4Gb explosion-proof certification to adapt to hazardous gas environments, humid water environments, and explosive gas environments underground. Specialized tools are also provided. Due to the numerous sensors, it can detect various parameters in multiple dimensions, and with the addition of a quantum gravimeter and a quantum magnetic field meter, its sensing capabilities are extremely sensitive.
[0111] In this embodiment of the invention, the mine inspection humanoid robot is 1.75 meters tall and weighs 2 kg. It can pass through almost 1% of the underground tunnels, has a maximum load of 3 kg, and can travel continuously for 1 km after a single hydrogen exchange.
[0112] The giant humanoid robot for mine inspection is 6.5 meters tall, weighs 5 tons, has a maximum load capacity of 4 tons, and can travel continuously for 2km on a single hydrogen refill.
[0113] like Figure 1 As shown, for initial fire suppression, the robot carries a dry powder fire extinguishing system, facilitating immediate initial extinguishing of any fire discovered at any time and place. This system references the piped dry powder fire extinguishing system used in building fire protection, and includes a dry powder storage container 608, a driving gas cylinder 609, a pressure reducing valve (located at the valve head of the driving gas cylinder 609, not shown), a driving gas cylinder 610, a solenoid valve 652, a nozzle 613, and a nozzle 614. The first five components of the system are housed in a box 615 carried by the robot, while the nozzle 613 extends from the box 615 and is strapped to the robot's right arm 618, and the nozzle 614 is located at the wrist.
[0114] like Figure 2As shown, in the visual sensor group, in addition to the camera array 621, depth camera 623, ultrasonic rangefinder array 624, and lidar 625 serving as SLAM navigation sensors, there is also an infrared thermal imaging 622 used to detect abnormal high-temperature points. When an abnormal temperature is detected, the robot approaches the location and uses the smoke sensor 634 and temperature sensor 639 to confirm again. If the video captured by the camera array 621 is added, the core processing unit 650 of the mine inspection humanoid robot judges the video. If it is identified by AI as flame or smoldering, it can be determined as the ignition point. The IMU inertial measurement unit 88, composed of accelerometers and gyroscopes, monitors the pitch angle and angular velocity of the robot's legs in real time. When the robot raises one leg, the IMU can instantly sense the body tilt and give millisecond-level instructions to the upper body to adjust the center of gravity and maintain overall balance. The foot sensor array 680 covers the entire foot. When the foot touches the ground with a step, the robot can know whether it has stepped off course or whether the force is uneven, so as to adjust its body posture in time.
[0115] While issuing a fire alarm via the wireless communication module 655, the core processing unit 650 opens the starting gas cylinder 610 via the solenoid valve 652. High-pressure air opens the driving gas cylinder 609 via the pipeline 659. The high-pressure nitrogen in the driving gas cylinder 609 enters the dry powder storage container 608, and the dry powder in it is sprayed out from the nozzle 614 through the nozzle 613. Since the nozzle 613 and the nozzle 614 are fixed on the robot's right arm, the robot only needs to point its right forearm 619 at the flames to extinguish the fire.
[0116] The UWB bidirectional mobile positioning base station 651 is essentially a combination of a personnel positioning card and a personnel positioning base station. It can receive signals from nearby personnel positioning cards and also send signals to nearby fixed personnel positioning base stations. In this way, the robot obtains the following information: ① The robot can travel to any mining face or dead corner of the tunnel, directly capturing signals from all surrounding active positioning cards over a wide area, automatically conducting area-wide roll call, headcount, personnel search, and blind spot personnel and vehicle statistics; ② After receiving the signal from the UWB bidirectional mobile positioning base station, the nearby fixed personnel positioning base station determines the robot's location and feeds it back to the robot via the underground high-speed communication network, thus providing the robot with its precise location information.
[0117] like Figure 1 , Figure 4As shown, in terms of inspection, initially multiple sensors were integrated into a small, automatically opening and closing container 663, which was carried by a robot using its dexterous hand and moved through the tunnel. This method worked well in small-section tunnels, but it had no advantage in large-section tunnels. Often, if the robot stood still, its arm could not reach the wind measurement point. If a wind measurement rod was used, the robot needed to move back and forth, affecting accuracy and efficiency, and making it impossible to dynamically and instantly grasp environmental parameters. After improvement, a telescopic rod was inserted into the robot's arm, which improved the problem somewhat, but it was still not suitable for taller and wider tunnels. Therefore, a custom-designed electric push rod 653 was installed on the robot's back, with a length set so that when the robot is in the center of the tunnel 604, its end can reach any point on the edge of the tunnel. The aforementioned automatically opening and closing container 663 is installed at the end. The base of the electric push rod 653 is divided into two parts: a fixed part 603 and a rotating part 601. The fixed part is connected to the robot, and the rotating part is connected to the telescopic rod, which can measure the required parameters at any point in the tunnel.
[0118] The sensors inside the automatic opening and closing container 663 are: a micro-pressure sensor 667, a smoke sensor 634, a carbon monoxide sensor 636, a carbon dioxide sensor 637, an oxygen sensor 673, a temperature sensor 639, an ultrasonic rangefinder 665, a dust concentration sensor 633, a gas concentration sensor 632, a wind direction sensor 631, a wind speed sensor 630, a laser rangefinder 660, a wireless charging battery 659, a support quality detection module 671, and a print head 672. In addition, the automatic opening and closing motor 661 can rotate under the control of the MCU to open or close the automatic opening and closing container 663.
[0119] like Figure 3 The image shows a remote control communication system for an underground robot based on quantum encrypted communication. The system consists of three levels: a ground control center 101, an underground 5G base station 102, and the robot body 103. The system is connected via optical fiber (long distance, several kilometers) and short-distance wireless transmission (50~2 meters) to achieve high-bandwidth, low-latency, anti-interference, and quantum-secure command and data interaction.
[0120] ① Ground Control Center 101: Operator 1 is responsible for sending motion instructions and control commands to the robot; 5G core network (including MPI / UPF) 201 processes control plane signaling and user plane data, and connects to external networks such as the AI cloud; quantum signal source 202 generates single-photon level quantum signals carrying encryption instructions and keys for quantum key distribution (QKD); 5G NR signal source 201 generates standard 5G New Radio signals (format can be CPRI / eCPRI or independent optical carrier), carrying high-bandwidth video streams, sensor data, etc. The above two signals (quantum signal and 5G NR signal) are transmitted over long distances (several kilometers) via optical fiber, and wavelength division multiplexing can be used to share the same optical fiber.
[0121] ② The underground 5G base station 102 completes the optical-to-wireless conversion and simultaneously transmits classical 5G signals and quantum microwave signals. This includes: ① Optical interface and signal separation: The quantum signal (single photon) transmitted from the optical fiber is fed into the quantum microwave converter (core conversion device) 204, converting the optical quantum signal into a microwave frequency quantum signal for easy wireless transmission; the 5G NR signal transmitted from the optical fiber is fed into the 5G baseband processing + RF front-end 203 to complete baseband processing and RF modulation, generating a standard 5G wireless signal; the 5G baseband processing + RF front-end 203 outputs a classic 5G signal (high bandwidth, low latency) for robot video backhaul and regular data communication; the quantum microwave converter 204 outputs a quantum microwave signal, which has anti-interference and low latency characteristics, used for transmitting encrypted commands and keys; ② Wireless transmission (distance 50~2 meters): The wireless link between the underground base station 102 and the robot body 103 operates within this distance, where strong electromagnetic interference may exist (such as large motors, frequency converters, cables, etc.); ③ Simultaneous transmission of two signals: The classic 5G signal has high bandwidth and low latency, suitable for short-distance point-to-point video / data transmission; the quantum microwave signal has anti-interference and low latency, used for core command and key transmission, ensuring secure communication under strong interference.
[0122] ④ Within the robot body 103: The robot body integrates equivalent receiving and processing modules; the 5G signal receiving module 205 receives classical 5G wireless signals emitted by the underground base station, and can also receive quantum microwave signals in the form of single photons / single waves; the quantum microwave converter 206 is symmetrical to the base station side, converting the received quantum microwave signals back into optical quantum or electrical signals for key extraction and command decryption; the local quantum key source 208 provides a one-time key for quantum key negotiation with the ground (such as the BB84 protocol), or for verifying / decrypting the received quantum key, realizing end-to-end quantum encryption; the robot core processing unit 650 is the robot's main control brain, receiving the operator's quantum-encrypted movement commands, processing uplink video and sensor data, and transmitting them back to the ground via the 5G uplink.
[0123] ⑤Signal flow summary: Downlink (ground → robot): Operator command → 5G core network → optical fiber (quantum signal + 5G NR signal) → underground base station → classic 5G signal + quantum microwave signal → wireless transmission (50~2 meters) → execution by robot core processing unit.
[0124] Uplink (robot → ground): Robot video / sensor data → 5G uplink wireless signal → underground base station → fiber optic → ground 5G core network → operator display / storage.
[0125] ⑥ Security Guarantee: The quantum signal carries the key via fiber optic and microwave relay, and works with the local quantum key source at the robot end to achieve quantum encrypted communication that is resistant to eavesdropping and interference. Especially in environments with strong electromagnetic interference, quantum microwaves ensure the reliable transmission of critical commands.
[0126] ⑦ Quantum Gravimeter 207: Utilizing the principle of atomic interference, an atomic cloud is cooled to near absolute zero using laser light and then released, allowing it to fall freely under gravity. At this point, the atoms exhibit quantum wave properties, and their wavefunction interference fringes shift due to extremely minute changes in gravitational acceleration. By measuring this shift, differences in underground density can be deduced, with an accuracy an order of magnitude higher than that of traditional gravimeters.
[0127] Core value: Quickly create 2D / 3D maps of underground resources without drilling, increasing detection speed by up to 10 times.
[0128] Strong anti-interference capability: The unique measurement mechanism can effectively overcome the vibration and environmental noise problems faced by traditional sensors.
[0129] Continuous and stable measurement: It can provide absolute gravity data that does not drift over long periods of time.
[0130] Miniaturization and ruggedness: The latest portable gravimeter has been shrunk to the size of a basketball and can withstand the operating environment of mobile platforms such as vehicles and robots.
[0131] ⑧ Quantum Magnetometer 207: Capturing Weak Magnetic Anomalies. Most underground minerals possess unique magnetic characteristics, and the quantum magnetometer is precisely the detector that captures these weak magnetic signals. Its sensitivity can reach the femtotes level (one hundred millionth of the Earth's magnetic field), capable of penetrating hundreds of meters of rock strata to detect anomalies.
[0132] The quantum gravimeter and quantum magnetometer are treated as a special high-order detection submodule, and they work together under the unified scheduling of the core processing unit.
[0133] ⑨ As the robot moves, the quantum gravimeter and quantum magnetometer act like a CT scanner scanning the ground, continuously providing unique field data that is distinct from all conventional sensors.
[0134] In Class I and II coal seam mines with a high tendency for spontaneous combustion, it is necessary to add ethylene, acetylene, and oxygen sensors; in non-coal mines without gas, gas sensors are not needed, but oxygen sensors are essential in radioactive mines, so the types of sensors can be changed as needed.
[0135] The "rotatable telescopic rod" essentially establishes a polar coordinate system on the measured tunnel cross-section. Similarly, a drilling rig was custom-designed for the robot, capable of drilling holes at any location within the tunnel to install various sensors and pipelines.
[0136] In addition, the robot's head is equipped with a built-in 360° mining lamp for illumination and external warning. Markers are placed at both ends of the roadway inspected by the robot and at the entrances of branch roadways to indicate the robot's operating range. The robot has been trained to possess underground inspection capabilities.
[0137] Deploying robots in the mine prevented three accidents within a year.
[0138] The first incident occurred when a pedestrian was passing through a connecting roadway and the ventilation door was not properly closed, causing a short circuit in the airflow. This reduced the air supply to the longwall face and caused the methane concentration to increase. The robot detected this during its inspection and issued an alarm. Based on the airflow and methane concentration records of the robot in that area, the operator quickly located the short circuit and promptly remotely controlled the robot to close the ventilation door.
[0139] The second incident involved a conveyor belt operator sleeping on duty, causing one of the belt's idlers to become stuck on a rock and rub violently against the belt, igniting coal dust. The robot followed the thick smoke to the location, quickly extinguished the fire, and sounded an alarm, waking the conveyor belt operator and stopping the conveyor belt, thus preventing a major fire accident.
[0140] The third instance occurred when the robot, while inspecting the tunnel, suddenly detected anomalies in gravity and magnetic fields: the gravitational acceleration *g* in a certain section of the tunnel was significantly lower than normal, while the magnetic field strength was abnormally high. The robot determined that the sudden increase in water mass in the goaf area was exerting an upward gravitational component on the quantum gravimeter, opposite to the direction of Earth's gravity, causing the measured gravitational acceleration value to decrease.
[0141] The filling of the goaf with water caused an abnormal increase in magnetic field strength detected by the quantum magnetic field instrument. The reasons are as follows: ① The water flow carries magnetic minerals. The rock strata around the goaf often contain magnetite and other minerals. The high-speed water flow washes away and transports these particles, causing them to accumulate locally and form a magnetic anomaly; ② When water flows through porous media or fractures, it generates a flow potential, which in turn induces a weak magnetic field (also known as electromagnetic coupling); ③ The rapid filling of a large amount of water may cause micro-fractures in the overlying rock strata, generating electromagnetic radiation, which is recorded as an enhanced magnetic field by the high-sensitivity quantum magnetic field instrument.
[0142] Based on this, the robot determined that water had accumulated in the goaf above and issued an emergency alarm. The mine immediately organized personnel to enter the site. Geophysical surveys were first conducted, strongly suggesting a large amount of water accumulation above. Therefore, decisive water-exploration measures were taken: a borehole was drilled upwards from the tunnel. After drilling more than 50 meters, a sudden gush of water gushed out of the borehole. The mine immediately sealed off the tunnel and activated pre-positioned drainage pumps to drain the water. This drainage operation lasted 16 days, removing a total of 150,000 cubic meters of water from above the goaf. If this water accumulation in the goaf had not been detected in time, it would have undoubtedly posed a significant safety hazard.
[0143] The mine owner proposed developing an application for a humanoid robot for mine inspection in mine pressure monitoring. After discussion, it was determined that the robot already possesses the basic hardware requirements for mine pressure monitoring. Firstly, utilizing SLAM+ underground personnel positioning system, decimeter-level positioning accuracy has been achieved, which basically meets the positioning requirements. However, adding some fixed positioning markers within the roadways would be even better (e.g., Figure 5 7 represents a lidar unit, and 381, 382, and 383 are white strip-shaped markers sprayed on the sidewalls of the tunnel. This system can correct the cumulative error of the SLAM+ underground personnel positioning system, achieving centimeter-level positioning accuracy. This lays the foundation for the uniqueness of cross-sections during repeated inspections, improving the reliability and accuracy of comparisons between multiple inspection results. Secondly, the lidar unit performs point cloud scanning on the tunnel cross-section, generating a 1:1 three-dimensional point cloud model. By comparing this model with historical inspection point cloud data, it can directly extract the roof subsidence, sidewall convergence, cross-sectional contour deformation, and three-dimensional displacement of local cracks, achieving millimeter-level absolute accuracy and generating dynamic 3D images. Furthermore, the precise dimensions of the cross-section can be measured by using a micro-pressure sensor at the end of the telescopic rod to measure along the tunnel contour.
[0144] The specific methods are as follows: (1) Using “SLAM + underground personnel positioning” or “SLAM + underground personnel positioning + fixed marker position correction in the roadway” or “SLAM + fixed marker position correction in the roadway” or “underground personnel positioning + fixed marker position correction in the roadway”, N measuring points are precisely arranged in the roadway; (2) At t0, the roadway segments corresponding to the first and second measuring points, the second and third measuring points, ... the N-1th to the Nth measuring points are precisely modeled in 3D (when scanning the 3D point cloud with laser radar, the robot must stand still at all N points and walk to the next point after scanning is completed). After all N-1 roadway segments are modeled, the 3D overall model of the entire roadway at t0 is constructed; (3) Similarly, the 3D overall model of each can be constructed at t1, t2, t3...tm; (4) The m 3D modeling diagrams are merged to form a dynamic 3D diagram of roadway deformation with time as the axis. Based on this diagram, the degree, trend, and rate of change of tunnel deformation can be monitored in real time, and the stability of the surrounding rock can be determined.
[0145] Alternatively, you can measure only the 2D cross-sectional view of the tunnel at each point, and follow the same steps as above. This method is slightly less accurate, but more intuitive.
[0146] In addition, the robot also achieves precise measurement of roadway deformation in the following ways: visual markers (signs, painted lines or characters, QR codes, barcodes) or electronic markers (passive radio frequency modules) are set at both ends of the inspected roadway and at the bends of the branch roadways, and / or detection markers are set at key detection locations to indicate the activity range; when monitoring roadway deformation using a telescopic rod, the measurement point is first determined by "SLAM + underground personnel positioning + fixed marker correction". If the floor is hard and free of debris, the floor is used as a reference, the telescopic rod extends upward, and the micro-pressure sensor at the top touches the roof and stops, recording the displacement. δ1, then δ2, so δ2-δ1 is the roof deformation. If the floor is soft or has debris, a ruler is sprayed or pasted on the tunnel side. The robot uses a laser line projector 669 to project a horizontal laser line onto the ruler and reads the reading C1. Then, the displacement δ1 of the telescopic rod is measured. Combined with the initial distance C from the laser line projector to the top of the automatic opening and closing container, the relative elevation and deformation of the roof are calculated. Measurements of other parts of the roof can be made by moving the robot or rotating the telescopic rod and measuring the slant length and angle, then deriving the results using trigonometric functions. When measuring the floor bulge, the telescopic rod extends downwards. When measuring the contraction of the two sides, the telescopic rod extends to both sides. In the above methods for measuring wind speed, wind direction, gas concentration, and tunnel deformation, the sensors are all encapsulated in the automatic opening and closing container 663 at the end of the wind measuring rod. An independent system with an MCU as the core completes data acquisition and wireless transmission.
[0147] The robot can also detect loose anchor bolts and inspect the quality of tunnel support.
[0148] The method for detecting loose anchor bolts in the roadway is to use an ultrasonic module. The ultrasonic probe is placed close to the end of the anchor bolt, and pulses are sent → echo amplification → filtering → AD sampling → original waveform upload. Then, the deformation is judged: a) Detecting noise in the middle of the thread contact surface: noise exists → gap exists → loose; b) Judging the equidistant nature of multiple echoes: uniform echo spacing → loose; disordered spacing → tightened under force; c) Identifying waveform distortion and widening: sharp and regular waveform → loose; widened and distorted waveform → tightened.
[0149] The system is used for quality inspection of tunnel support, including measuring tunnel height, width, over-excavation, and under-excavation; measuring shotcrete thickness using ultrasonic and / or electromagnetic waves; taking and uploading photos of the metal mesh installation; measuring anchor bolt density and spacing; measuring anchor bolt tightness; and measuring shotcrete thickness, masonry thickness, and grouting quality. The support quality inspection module 671 can also be used to test roof delamination, shotcrete support thickness, masonry thickness, and grouting quality. Its integrated ultrasonic, electromagnetic, and stress wave detection module 671 can complete a series of processes, including pulse transmission, echo amplification, filtering, AD sampling, AI deep analysis, and result output.
[0150] Another task of the robot is to photograph, identify, and upload images of the display interfaces of various instruments, equipment, and switches within a designated area, and to detect the temperature of their casings and related cables (using infrared thermal imagers or temperature sensors integrated into fingers), as well as the presence of odors (using machine olfactory sensors). For this purpose, a dual-spectrum camera 643, model FLIR lepton SDS, is installed on the robot's head to photograph the instruments above and monitor the temperature of the cables and electrical components. The robot can also automatically read historical data stored internally in various instruments via near-field communication (NFC / RFID).
[0151] Another task for robots is to perform simple, repetitive tasks, such as moving materials. Figure 6 Toolbox 388 needs to be moved forward underground as the working face advances. After loading and training the robot assembly and disassembly model, it can be moved.
[0152] With the development of predictive AI models, the deep learning models loaded in the core processing unit of robots will be able to predict the risk of gas outburst and the probability of roof collapse in the next 24 hours based on continuously collected time-series data such as wind speed, gas outburst, and roadway deformation rate, and proactively propose intervention measures such as emission measures and strengthening support.
[0153] Example 2 This invention provides a humanoid mining robot for replacing workers in hazardous jobs at the mining face. The replaced workers can then remotely control the robot from the ground. This robot eliminates the need for sensors, fire extinguishers, and other unnecessary equipment in mining operations. This invention manufactures three underground mining robots, named "Mining Robot No. 1," "Mining Robot No. 2," and "Mining Robot No. 3." These are general-purpose underground mining robots; different working models can be loaded onto them to complete corresponding tasks.
[0154] In this embodiment of the invention, three mining robots are first used to initially replace the coal release workers at the fully mechanized longwall face. The reason is: ① Highest risk: Coal dumping operations face the risk of roof coal collapse, coal slag, roof spalling, gas outbursts, and dust. It is a high-risk position for casualties and serious injuries, and personnel are in a dynamic danger zone for a long time. ② Extremely poor working environment: high dust, high gas, poor visibility, narrow space, long hours of manual labor with high intensity and high risk of occupational disease; ③Standardized and easily automated: The coal feeding process is fixed, with clear rules for opening and closing the coal feed port, observing the coal flow, stopping when ore is encountered, and cyclical operation. Robot vision + automatic control can easily replace it, without the need for complex on-site judgment.
[0155] The first experiment yielded good results.
[0156] Three top coal caving robots in the fully mechanized top coal caving face are arranged in an orderly manner along the hydraulic support column and work together in a fully automated manner to accurately complete processes such as coal caving, quantity control, and support protection.
[0157] like Figure 7 As shown, mining robot No. 1 is responsible for monitoring the working condition and controlling the opening and closing of the support. It is closely attached to the control end of the hydraulic support and uses vision, pressure and material flow sensors to monitor the roof pressure, support posture and coal flow status in the mining area in real time. According to the coal feeding step distance and sequence, it remotely controls the opening and closing of the coal feeding port of the corresponding support to control the coal feeding rhythm. When it detects large pieces of coal getting stuck or excessive coal feeding, it immediately issues an early warning and finely adjusts the opening width.
[0158] Mining Robot No. 2 is dedicated to coal flow sorting and clearing. Positioned below the coal discharge port, its robotic arm operates flexibly, crushing and clearing large pieces of coal and loose stones stuck in the chute and discharge port. Simultaneously, it distinguishes between coal and loose stones, strictly controlling the loose stone content to prevent large amounts of loose stones from entering the conveyor, ensuring the quality of raw coal and smooth transportation.
[0159] Mining Robot No. 3 is responsible for surrounding inspections and safety protection. It moves back and forth along the gaps between the supports, inspecting the roof, roadway sides, and equipment connections for potential hazards such as roof collapse, falling roofs, and support tilting, while simultaneously monitoring dust and gas concentrations. Safety warning zones are set up at intervals in the coal discharge area to prevent coal flow from impacting and damaging the equipment. After the operation is completed, the closure of each coal discharge port is checked to confirm that the site is safe.
[0160] The three robots exchange data in real time through industrial communication, strictly adhering to the segmented, rotating, and uniform coal release process, maintaining a safe operating distance from each other, and ensuring seamless workflow. The entire operation is unmanned, continuously cooperating with the coal mining system to efficiently complete the entire top coal caving process.
[0161] In this embodiment of the invention, five additional mining robots are customized for the tunneling face, named "Mining Robot No. 4" to "Mining Robot No. 8". Mining Robot No. 4 serves as the tunneling operator, Mining Robots No. 5, 6, and 7 are support workers, and Mining Robot No. 8, equipped with a quantum gravimeter, serves as the blasting operator. Their working process is as follows: The underground tunnel is well-organized, with five mining robots working in a coordinated manner, operating according to preset programs and sensor data, without any human intervention on site.
[0162] like Figure 8As shown, the No. 4 mining robot sits inside the control cabin of the roadheader, precisely controlling the cutting arm to swing up and down and left and right, cutting the coal and rock mass at a uniform speed, and monitoring the cutting resistance, machine vibration, and roadway profile in real time. When encountering hard rock or grit, it automatically reduces the cutting speed and simultaneously sends a working condition warning to the other robots; the coal grit produced by cutting is transported outward by the scraper conveyor, and the robot continuously controls the tunneling progress and cross-sectional dimensions.
[0163] like Figure 9 As shown, mining robots 5, 6, and 7 follow closely behind the roadheader, working in separate zones to coordinate and complete the roadway support. One robot is responsible for moving the temporary support frame forward, precisely aligning it with the roof, and monitoring roof delamination and surrounding rock pressure in real time to ensure proper support. The other two robots are responsible for installing anchor bolts, anchor cables, and steel strips. The robotic arms are precisely positioned and automatically complete drilling, material feeding, and tightening processes, systematically arranging support components along the roadway sides and roof. The three robots maintain a safe working distance, transmit position signals to each other, and move forward synchronously with the tunneling progress to solidify the surrounding rock protection of the roadway.
[0164] like Figure 10 As shown, when mining robots 5, 6, and 7 were drilling anchor bolts underground, three operators on the ground were remotely controlling them in real time.
[0165] After the support was completed and the site conditions were safe, mining robot No. 8 entered the hard rock section that the roadheader could not handle. With the assistance of mining robots No. 4, 5, 6, and 7, it completed the blasting operation. The entire process adhered to underground blasting safety regulations. First, a "one-blast, three-check" procedure was performed, using gas sensors to detect methane and coal dust concentrations. Once the levels were confirmed to be within acceptable limits, shallow-hole loosening blasting operations were carried out. Holes were precisely drilled according to the marked locations, and a small amount of explosives and detonation devices were loaded. After the operation was completed, the team withdrew to the designated safe area. Once all personnel and equipment had evacuated and the perimeter was secured, the blast was remotely detonated according to instructions. After the blast, the smoke was allowed to dissipate, and the environmental safety was checked again before a signal indicating the blast was complete was sent to the entire team.
[0166] During one blasting operation, after loading explosives at the blasting face, Mining Robot No. 8 pulled the detonation cord outwards. Shortly after moving forward, it received a signal from ground-based remote control to initiate the blast. Using its onboard quantum gravimeter, Mining Robot No. 8 obtained its precise coordinates. Calculating the distance between its current location and the blasting face using these coordinates, it determined that the distance was only about 30 meters. Therefore, Mining Robot No. 8 initiated a denial procedure. It only carried out the blast after pulling the detonation cord to a position more than 1 meter away and conducting a complete safety check.
[0167] The overall operation process is interconnected: tunneling and cutting → follow-up support → local loosening blasting (activated when encountering hard rock / structures). The five robots are interconnected by the underground industrial network, automatically avoid operational interference, strictly follow the safety regulations of the tunneling face, and carry out tunneling operations continuously and stably.
[0168] Example 3 This invention provides a giant humanoid robot for mine inspection. This robot is used to survey and map the gravitational and magnetic field gradients of 112 coal mines under its jurisdiction. The total area of these 112 coal mines is 1826 square kilometers. Mapping them using a robot that walks automatically would take several years. To speed up the process, a combination of drones and robots was used for surveying. Initially, a cage was installed on the drone 61, and later a ring-shaped sling was used for lifting, but the results were unsatisfactory. Finally, a lifting ring 64 was added to the top of the robot, and a corresponding automatic fast-locking and detaching lifting device was added to the drone 61. One end of the lifting device is connected to the lifting ring 64, and the other end is connected to a sling 62, which is then connected to the drone 61. This allows the drone 61 to lift the robot 63 to any location very quickly, increasing the survey speed by more than 10 times and saving the robot 63 a significant amount of time spent climbing mountains and wading through water. A mine can be surveyed in approximately three to four days.
[0169] When the robot was hoisted to the designated survey area of Liushenyu Coal Mine, the UAV 61 automatically located the preset survey benchmark point through airborne RTK positioning and lidar ground scanning. After reaching the point directly above it, the UAV slowly descended to a height of about 2.5 meters above the ground, and the sling 62 was also slowly retracted. Then the robot's legs were in a slightly crouching position, which served as a cushion when the feet touched the ground. The robot sent a "stable ground contact" signal to the UAV, and the automatic quick-locking mechanism on the hoist immediately unlocked, the hook separated from the lifting ring 64 on the top of the robot, and the UAV then flew away with the sling 62 to a nearby open area for rest. After the robot stood independently, it used its built-in inertial measurement unit (IMU) and inertial stabilization platform to perform self-leveling, eliminating tilt and vibration interference.
[0170] Next, the robot began measuring the gravitational and magnetic field gradients at that point. The gradient-sensitive head of the quantum gravimeter extended downwards from its abdomen to measure the gravitational acceleration *g*. After the gravitational acceleration *g* was measured at each measurement point on the grid node, a contour map of the gravitational acceleration within the mine area appeared on the mining engineering plan. The quantum magnetometer operated in SERF (spin-free exchange relaxation) mode, simultaneously acquiring the spatial rate of change of magnetic induction intensity in the X, Y, and Z directions through four orthogonally arranged magnetic field gradient probes. Similarly, a contour map of magnetic field intensity in the three directions could be generated. Each measurement point was measured five times, with an integration time of 30 seconds each time. The data was automatically filtered to remove residual flight vibrations and geomagnetic pulsation noise, and transmitted back to the command center in real time. After completing the measurement at each point, the robot autonomously planned the coordinates of the next point and issued a new hoisting request to the drone.
[0171] During a point-by-point survey of the Liushenyu Coal Mine, the robot discovered a strong negative anomaly in the vertical gravity gradient deep in the northeast direction of the mine, 120 ehrls lower than the surrounding background value. Simultaneously, the magnetic field gradient exhibited irregular, spike-like disturbances, clearly inconsistent with the characteristics of a normal coal seam goaf. The robot immediately refined the measurement point grid in this area and used the deep inversion algorithm of the quantum gravimeter for imaging. The results clearly revealed a mining space never before marked on the mine's blueprints on the three-dimensional gravity gradient profile. The lower gravitational acceleration was due to the presence of a goaf below, while the irregular, spike-like disturbances in the magnetic field gradient indicated the presence of a large number of steel mining machines in the "dark side" deep underground. The operator used a drone to hoist the robot to the entrance of the mine's auxiliary inclined shaft. The robot, programmed to ride various mine vehicles, took a "monkey car" (an aerial personnel carrier) to the bottom of the auxiliary inclined shaft, and then a level roadway personnel carrier to a temporary shunting yard in the middle of the main roadway on the northeast wing of the minefield. Using measurements of gravitational acceleration and magnetic field gradient anomalies on the ground, the robot determined that the "dark side" was to its left and slightly ahead in the direction of travel in the main roadway. The robot continued forward, reaching the end of the roadway where a downhill roadway, not marked on the "Mining Engineering Plan," appeared. With the 5G signal nearly cut off, the operator sent a command to the robot to continue exploring without remote control. The robot continued downhill for nearly 5 meters, then turned right for 1 meter before reaching the "dark side" located in the lower coal seam (this working face was outside the scope of the "Mining License," constituting illegal mining beyond the permitted layers). Nearly 30 people were working at the working face, but none of them were carrying personnel positioning cards. At this moment, the gas concentration sensor carried by the robot issued a warning: the gas concentration at the working face had reached 2.1%, severely exceeding the evacuation and power-off concentration standard (1.5%). However, this "dark side" had no gas sensor installed, nor was it staffed with a safety officer. The robot immediately issued a warning through its loudspeaker: "The gas concentration at the working face is severely exceeded. Please evacuate immediately and cut off the power!" Some people wanted to evacuate after learning of the situation, but the team leader threatened them with not receiving their wages, causing everyone to continue working as usual. So the robot immediately returned along the same route, and when a 5G signal appeared, it immediately transmitted the video, photos, and monitoring data it had just taken to the network. After learning of the situation, the management immediately organized personnel to go to Liushenyu Coal Mine and sent people down the mine to stop the continued operation in the "dark side," cutting off the power, evacuating personnel, and sealing off the area, thus averting a major gas explosion accident.
[0172] Giant humanoid robots are used for mine inspections. They are tasked with surveying the mine surfaces in mountainous and densely forested areas with steep terrain. The slopes in these areas exceed 45 degrees, making it impossible for even drones and ordinary robots to stand upright, let alone carry out measurement work.
[0173] exist Figure 11In the robot, operator 1, acting as the driver, sits in chair 2. A camera 3 is positioned in front of operator 1 to capture their movements, which are then mapped onto the robot's limbs. The entire cockpit 4 is a sealed container. Operator 1 is positioned in front of the cockpit 4 through infrared-resistant hollow glass 5. Operator 1 also wears infrared thermal imaging glasses 6, which can detect flames within dense smoke. A lidar 7 is installed on the top of the cockpit 4. The robot's neck 8 serves as a pitch axis, allowing for tilting movements. The neck 8 and limbs are connected to the torso 10 via joints. A remote control 9 serves as the human-machine interface, enabling both human-machine interaction and remote control of the robot's operations.
[0174] The shoulder joint consists of a left shoulder joint pitch axis 11, a right shoulder joint pitch axis 12, a left shoulder joint yaw axis 13, and a right shoulder joint yaw axis 14. The left upper arm 15 and right upper arm 17 are also connected to a left upper arm roll axis 18 and a right upper arm roll axis 19, respectively. The left and right elbow joints are connected to a left elbow joint pitch axis 20 and a right elbow joint pitch axis 21, respectively. A turret 24 is connected to the left wrist 22 and right wrist 23, respectively. The center of the turret is connected to the robot's forearms 26 and 27. The turret 24 is connected to three tools at 120° angles in the radial direction. When a pair of tools is needed, the turret is rotated to the corresponding position. These three pairs of tools are: Group A: wheels 29 and 30; Group B: hydraulic calipers 31 and 32; and Group C: spiked shoes 33 and 34. When a certain tool is needed, only the corresponding tool needs to be rotated to the required position; its principle is similar to the automatic turret of a CNC machine tool. The hip joint is composed of the left hip joint pitch axis 51, the right hip joint pitch axis 52, the left hip joint roll axis 53, and the right hip joint roll axis 54. The knee joint is composed of the left knee joint pitch axis 61 and the right knee joint pitch axis 62. The ankle is composed of the left ankle pitch axis 71 and the right ankle pitch axis 72.
[0175] All of the above-mentioned joint axes are driven by joint-type drive devices such as electric RV reducers, hydraulically driven swing cylinders, and hydraulic piston cylinders.
[0176] The ankle is equipped with a left track walking mechanism 81 and a right track walking mechanism 82, which can move forward by stepping with the legs or by rotating the tracks without moving the legs, giving it excellent off-road capability.
[0177] Camera 3 captures various movements of operator 1's right arm and transmits them to the core processing unit 650, which consists of RK3588+STM32. After deconstruction, the images are mapped onto the head, neck, and limb power joints 99, which drive the robot to walk.
[0178] like Figure 13When the robot moves in a quadrupedal posture, the turret 24 rotates the spikes 33 and 34 to a position perpendicular to the ground; for example... Figure 14 When the robot moves in a four-wheeled configuration, the turret 24 rotates wheels 29 and 30 to a position perpendicular to the ground; for example... Figure 15 When the robot climbs in an upright position, the turret 24 rotates the hydraulic clamp 32 to the front, which is used as an auxiliary tool for climbing steep terrain. It is clamped to a reliable tree trunk or branch to assist bipedal climbing, and the forearm 15 is a telescopic structure.
[0179] In addition, the operator can issue voice commands via microphone 37, or send digital signals via joystick 38 and integrated helmet system 39 to the core processing unit 650. The core processing unit 650 manipulates the head, neck, and limb power joints 506 to complete the movement according to the commands or digital signals. The power comes from the battery pack 656 composed of fuel cells and supercapacitors or the hydraulic pump 93. The control signals come from the core processing unit 650, which includes an electric drive RV reducer, a hydraulic drive swing cylinder group, a hydraulic piston cylinder group, track drives 81 and 82, and hydraulic calipers 31 and 32.
[0180] The core processing unit 650 is also connected to an environmental perception module, including an infrared thermal imager 622, an inertial measurement unit 88, a lidar 625, and a 4D millimeter-wave radar 626. The infrared thermal imager 622 is used to identify flames in dense smoke, aiding in precise fire suppression. The lidar 625 scans the surrounding environment by emitting laser beams, quickly generating a high-precision 3D point cloud map, enabling precise ranging, positioning, identification, and tracking of targets such as obstacles, lanes, and pedestrians. It is a key component of SLAM navigation. The 4D millimeter-wave radar 626 is... An important supplement to the lidar, it can detect targets at long distances and measure the speed of dynamic targets in rain, snow, and dust environments; a pair of LM3mini camera rangefinders 628 are also installed on the outside of the lower leg, which can assist in SLAM navigation when the lidar and 4D millimeter-wave radar lines of sight are obstructed; the inertial measurement unit (IMU) 88 is composed of accelerometers and gyroscopes, which monitors the pitch angle and angular velocity of the robot's legs in real time. When the robot raises one leg, the IMU can instantly sense the body tilt and give millisecond-level instructions to the upper body to adjust the center of gravity and maintain overall balance.
[0181] like Figure 15 The caliper 32 can also be used as an auxiliary tool when climbing steep terrain, by clamping it onto a reliable tree trunk or branch to assist bipedal climbing.
[0182] In addition, the operator can issue voice commands via microphone 37, or send digital signals via joystick 38 and integrated helmet system 39 to the core processing unit 650. The core processing unit 650 manipulates the head, neck, and limb power joints 506 to complete the movement according to the commands or digital signals. The power comes from the battery pack 656 composed of fuel cells and supercapacitors or the hydraulic pump 93. The control signals come from the core processing unit 650, which includes an electric drive RV reducer, a hydraulic drive swing cylinder group, a hydraulic piston cylinder group, track drives 81 and 82, and hydraulic calipers 31 and 32.
[0183] The core processing unit 650 is also connected to an environmental perception module, including an infrared thermal imager 622, an inertial measurement unit 88, a lidar 625, and a 4D millimeter-wave radar 626. The infrared thermal imager 622 is used to identify flames in dense smoke, aiding in precise fire suppression. The lidar 625 scans the surrounding environment by emitting laser beams, quickly generating a high-precision 3D point cloud map, enabling precise ranging, positioning, identification, and tracking of targets such as obstacles, lanes, and pedestrians. It is a key component of SLAM navigation. The 4D millimeter-wave radar 626 is... An important supplement to the lidar, it can detect targets at long distances and measure the speed of dynamic targets in rain, snow, and dust environments; a pair of LM3mini camera rangefinders 628 are also installed on the outside of the lower leg, which can assist in SLAM navigation when the lidar and 4D millimeter-wave radar lines of sight are obstructed; the inertial measurement unit (IMU) 88 is composed of accelerometers and gyroscopes, which monitors the pitch angle and angular velocity of the robot's legs in real time. When the robot raises one leg, the IMU can instantly sense the body tilt and give millisecond-level instructions to the upper body to adjust the center of gravity and maintain overall balance.
[0184] When the giant humanoid robot is operating in the mine, it uses a four-wheeled mode on gentle slopes, and an upright mode on steps and steeper areas, even climbing steps as high as 1.5 meters. When there are few obstacles, it uses an upright wheeled mode. When the slope is between 30° and 45°, it uses a four-legged mode. The four-legged mode provides strong traction because the two front legs are spiked and the two hind legs are tracks. When the slope exceeds 45°, it walks like a human, using an upright mode with two legs as the main support and the arms as clamps to climb tree trunks and large branches. When the slope is even steeper, the co-pilot first hooks the winch onto a tree trunk, and then uses a four-wheeled mode plus the winch to pass through. When encountering large cliffs, it uses a detour tactic to go around them. If there are trees blocking the way and it is impossible to go around them, it takes out a giant chainsaw to quickly cut them down. Regardless of the posture, the robot always keeps the driver's seat facing the direction of travel through the positioning of the built-in gyroscope. During the mountain climb, the robot passed through many bushes, obstructing the line of sight of the lidar and 4D millimeter-wave radar. The robot quickly switched the navigation mode to be dominated by a pair of LM3mini miniature camera rangefinders installed on the outside of the calf.
[0185] Example 4 This invention provides a humanoid robot for mine fire rescue. Mine fire rescue is an important component of mine safety.
[0186] like Figure 16 As shown, the humanoid robot for mine fire rescue is covered with three protective layers: the outermost layer (55) is carbon fiber, the middle layer (56) is hot melt adhesive, and the innermost layer (57) is a cooling pipe network. The carbon fiber layer, with its high strength, high temperature resistance, and non-combustible properties, serves as the outermost high-strength protective layer. The middle layer, hot melt adhesive, has extremely high thermal insulation properties and is the lightest and best-insulating material discovered by humankind. The innermost cooling water pipes carry away any residual heat leaked from the upper two layers. The cooling water in the cooling pipe network comes from firefighting water.
[0187] like Figure 17 The main battery 657 is a hydrogen fuel cell, and the backup battery is an aluminum-air battery. The robot's shoulder joint can rotate more than 360° and its hip joint can rotate more than 225°, making it suitable for working conditions such as passing through narrow gaps and getting up and getting out of trouble after falling.
[0188] like Figure 18 The lifting lug 361 is used for lifting the aircraft. The hydraulic shears 371 and angle grinder 372 are directly mounted on the robot's arms for easy obstacle removal during firefighting. First, the safety pin of the fire extinguisher is pulled out, and the automatic presser 374 is pressed onto the fire extinguisher handle 373. The robot only needs to give a signal to press the handle 373 for the fire extinguisher 375 to spray, without manual operation. The water hose or fire extinguisher nozzle 376 and the nozzle are quickly secured to the nozzle universal clamping mechanism 379 by a quick-release clip 377, allowing the nozzle 378 to spray at any angle of fire.
[0189] like Figure 19The hydraulic shears 371, angle grinder 372, automatic presser 374, X-axis universal clamping motor 385, Y-axis universal clamping motor 386, booster pump 236, and wireless communication module 655 are all controlled by the core processing unit 380 of the general-purpose fire-fighting humanoid robot. The X-axis universal clamping motor 385 and Y-axis universal clamping motor 386 allow the nozzle 378 to spray at any angle forward. The booster pump 236 pressurizes tap water to achieve atomization pressure, increasing the fire extinguishing effect. 621 is a camera array, 622 is an infrared thermal imaging camera, 623 is a depth camera, 624 is an ultrasonic rangefinder array, and 625 is a lidar as the core components for robot SLAM positioning. The 4D millimeter-wave radar 626 is an important supplement to the lidar, enabling long-distance detection and dynamic positioning in rain, snow, dust, and dense smoke environments. Target velocity measurement; the IMU (Inertial Measurement Unit) 638, composed of an accelerometer and a gyroscope, monitors the pitch angle and angular velocity of the robot's legs in real time. When one leg of the robot lifts, the IMU can instantly sense the body tilt and provide millisecond-level commands to adjust the upper body's center of gravity to maintain overall balance; the gas sensor 632, carbon monoxide sensor 636, carbon dioxide sensor 637, hydrogen sulfide sensor 390, and temperature sensor 639 are essential for detecting whether there is smoke or fire in the detection area; the life detector 382 is sensitive and reliable when searching for trapped people; the satellite navigation and positioning system 84 is a basic navigation configuration; the battery 656 adopts a quick-replacement structure to not delay firefighting; the wind speed sensor 630, wind direction sensor 631, and temperature sensor 632 are used to monitor changes in wind speed and direction, and the temperature of the fire scene.
[0190] The humanoid robot for mine fire rescue is equipped with impact-resistant cushioning armor on key body parts (head, torso, joints), preventing injury from collapses and impacts. If the force of lifting one leg reaches 590N, the self-rescue mode is triggered. The robot weighs 2kg (1960N), and a normal leg lift requires 392N. If the lifting force exceeds 392N × 1.3 = 509.6N, it is considered that the leg is stuck and the self-rescue mode is activated: both arms and the other leg work together to help the stuck leg free.
[0191] Finally, three points should be made: First, the above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention, but it should not be considered that the specific implementation of the present invention is limited to these descriptions. Second, the robots in the embodiments each have their own focus, but it cannot be said that a certain technical feature is limited to the application scenarios of that embodiment. Under certain conditions, a certain embodiment and its corresponding claims can be applied to other embodiments and claims. Third, for those skilled in the art, several simple deductions or substitutions can be made without departing from the concept of the present invention, and all of these should be considered to fall within the protection scope of the present invention.
Claims
1. A humanoid robot for mining, comprising: The core processing unit, mobile module, environmental perception module, job execution module, environmental monitoring module, protection module, power module, positioning and navigation module, and communication module are characterized in that: The core processing unit is electrically connected and / or signal connected to the mobile module, environmental perception module, environmental monitoring module, power module, positioning and navigation module, and communication module, respectively. The outer casing is made of non-combustible or flame-retardant materials, and the overall protection rating is at least IPX4; The mobile module has pre-stored multi-pose switching programs and programs for navigating various complex terrains; Furthermore, the robot satisfies at least one of the following technical features: (A) The power module is a fuel cell; (B) The load-bearing joint of the power module adopts a hydraulically driven swing cylinder or a combination of a hydraulically driven swing cylinder and a geared motor; (C) The environmental perception module includes a gravimeter and / or a magnetometer, which provides the robot with absolute position and geological perception signals that are independent of the navigation system and immune to electromagnetic interference; (D) The protection module also includes a heat-absorbing layer located in the outer shell, which includes a cooling pipe network that uses fire-fighting water as cooling water; and the communication module employs encryption measures.
2. The mining humanoid robot according to claim 1, characterized in that, It also includes at least one of the following items: (1) The environmental monitoring module further includes a dust concentration sensor and / or a combustible gas sensor; (2) The environmental perception module integrates one or more of the following sensors: visible light camera array, rangefinder array, lidar, laser rangefinder, depth camera, 4D millimeter-wave radar, infrared thermal imaging module, foot pressure sensor array, inertial measurement unit, and a visible light and infrared dual-spectrum camera with the lens pointing directly upward. (3) The mobile module and the operation execution module support autonomous control and manual operation. The human-based operation mode uses motion capture, electromyography signals or brain-computer interface to map information to the corresponding body parts of the robot, or completes the operation through helmet system, voice, and joystick. During movement, the position and video are transmitted back to the operation terminal, and a real-time operation map is built based on SLAM to achieve collaborative planning. The human-assisted operation mode indicates the direction and task to the operator, and the robot acts autonomously. (4) The environmental perception module also includes cameras and / or ranging sensors distributed on the legs and / or arms. When the field of view at the top of the body is blocked, the core processing unit takes over the navigation control from the sensors through visual servoing and corrects the body movement trajectory in real time. (5) The complex terrain is supported by the program to switch between side-lying, crouching, upright, quadrupedal, four-wheeled, climbing, riding, hoisting and winch modes; wherein, the climbing action is completed by the arm in conjunction with the end effector. The arm is of fixed length or telescopic structure, and the end effector is any one of the following: dexterous hand, caliper, gripper, octopus soft claw, hook, bucket or other reliable object that can be firmly grasped in front. After the telescopic arm is extended, it can grasp a fixed support at a preset distance. The aforementioned "riding" refers to the robot mimicking workers riding in underground personnel-carrying vehicles and devices. The aforementioned hoisting refers to the use of drones to lift objects to the target location while the robot is operating on the ground. The winch mentioned above is a device that allows the robot to climb steep terrain or escape from trouble, mimicking the use of an off-road vehicle. The winch is fixed to the robot's shoulder or torso. (6) The program also includes redundancy of degrees of freedom in accordance with the complex terrain, that is, the shoulder joint can rotate at an angle greater than 360° and the hip joint can rotate at an angle greater than 225°. (7) The protection module also includes an explosion-proof design, and the core electrical components are encapsulated in an explosion-proof manner; (8) The gravimeter is a quantum gravimeter, and the magnetic field meter is a quantum magnetic field meter; (9) The control system has a built-in multi-posture switching program, which can freely switch between upright humanoid, four-legged bionic, four-wheeled driving, and crouching low center of gravity two-wheel posture.
3. The mining humanoid robot according to claim 1, characterized in that, The tasks that the robot needs to perform include one or more of the following: replacement, inspection, remote sensing, measurement and control, quality inspection, mine pressure monitoring, meter reading, temperature measurement, odor detection, fire extinguishing, relocation, and tagging. - The replacement refers to robots replacing workers in underground work; - The aforementioned inspection involves the robot using a UWB bidirectional mobile positioning base station to capture signals from surrounding active positioning cards, enabling roll call, headcount, and blind spot personnel statistics, as well as recording and correcting violations; - The remote sensing is used by the robot to carry the gravimeter and / or magnetometer to detect the mining face or goaf, to prevent over-excavation or to detect water hazards in advance. - The measurement and control refers to mobile wind measurement and mobile monitoring, which measures wind speed, wind direction, and gas concentration at any cross section underground, or measures toxic and harmful gases in roadways that have not been ventilated for a long time. - The quality inspection includes one or more of the following: measuring the dimensions of the tunnel section, taking photos to record the quality of the metal mesh laying, measuring the anchoring status of the anchor bolts, the thickness of the shotcrete, the thickness of the masonry arch, and the quality of the grouting. - The mining pressure refers to the measurement of the roof delamination, sidewall shrinkage, and floor bulge in the roadway. - The reading of the meter can be achieved by taking a picture of the display interface of the meter, equipment, or switch in the specified area, identifying and uploading the image, or by reading data using an NFC module; - The temperature measurement refers to detecting the temperature of the electrical equipment casing, cables, and junction boxes; - The odor refers to the abnormal gas emitted by a local fire or spontaneous combustion of coal seam, detected using a machine olfactory sensor or harmful gas sensor. - The fire extinguishing refers to immediately alarming, cutting off power, and extinguishing the initial fire upon discovering a flame or signs of fire; - The relocation refers to the dismantling, moving, and reassembling of the monitoring and control equipment at the mining face; - The mark is a slogan or mark printed at a designated location by inkjet or printing.
4. The mining humanoid robot according to claim 1, characterized in that, It also includes one or a combination of the following items: (1) The environmental monitoring module also includes a sensor group, which includes a rangefinder, wind speed sensor, wind direction sensor, smoke sensor, carbon dioxide sensor, radon sensor, machine olfaction sensor, oxygen sensor, micro-pressure sensor, temperature sensor, humidity sensor, support quality detection module, and inkjet head, or multiple modules operating in conjunction; the support quality detection module is equipped with a detection module including an ultrasonic probe, which can detect one or more parameters in anchor bolt loosening, top plate delamination, and support thickness; (2) Drive redundancy control for the mobile module and the environmental perception module: When the force load on a single leg exceeds the preset threshold, the multi-limb collaborative escape algorithm is automatically triggered, and the stuck limb is released by the traction force of the remaining limbs. (3) The power module also includes a supercapacitor as a power buffer for the fuel cell; (4) The fuel cell is a hydrogen fuel cell, and the hydrogen storage method is solid-state hydrogen storage; (5) The sensor group is fixed to the end of a telescopic rod. The sensor group at the top of the telescopic rod is an independent sensor system centered on the MCU. Each sensor establishes communication and electrical connection with the MCU, and the MCU establishes communication with the core processing unit. The telescopic rod is divided into a standard type and a rotating type. The rotating type has a base consisting of a fixed part and a rotating part. The fixed part is connected to the robot, and the rotating part is connected to the telescopic rod, which can measure various parameters of any cross section in the tunnel at the location. The standard type requires the robot to move laterally, and then coordinates with the extension and retraction of the telescopic rod to make measurements.
5. The mining humanoid robot according to claim 3 or 4, characterized in that: The aforementioned quality inspection and mine pressure also include using the environmental sensing module or telescopic rod to perform precise 2D modeling of the cross-sectional dimensions of the roadway corresponding to a certain measurement point underground, as well as the deformation over time; The steps for measuring tunnel dimensions and deformation using an environmental sensing module are as follows: The robot arrives at the measurement point and is stationary. It collects point cloud data of the tunnel cross section through LiDAR scanning, filters and reduces noise of the point cloud data, corrects deviations, extracts the tunnel outline and fits the cross section model, and calculates the net dimensions of the tunnel. The steps for measuring the cross-sectional dimensions and deformation of a roadway using a telescopic rod are as follows: ① Use "SLAM + underground personnel positioning system" or "SLAM + underground personnel positioning + fixed marker position correction in roadway" to determine the measurement points; ② If the base plate is hard and free of debris, the elevation of the base plate will not change during each measurement and can be used as a measurement reference point; When the telescopic rod extends upward, the micro-pressure sensor at the top touches the tunnel roof and is triggered. The MCU receives the signal and immediately sends a signal to the robot's wireless communication module 655 through the sensor group's wireless communication module 658. The robot's core processing unit immediately stops the electric push rod and receives its displacement δ1. Assuming the next displacement is δ2, then δ2-δ1 is the deformation of the roof. ③ If the base plate has low hardness and contains debris, it cannot be used as a measurement benchmark. A ruler needs to be sprayed or pasted on any side of the roadway. During measurement, the robot adjusts the laser line projector to be horizontal and turns it on. The laser horizontal line is irradiated onto the ruler. The robot uses a visible light camera to capture and identify the reading C1 of the laser line on the ruler. Then, the electric push rod is turned on and the displacement δ1 of the electric push rod is measured according to the method in ②. The relative elevation of the roadway roof is C1+δ1+C, where C is the distance from the laser line projector to the top of the automatic opening and closing container (before the telescopic rod is extended). The result of the next measurement is C2+δ2+C. Then (C2+δ2+C)-(C1+δ1+C)=C2-C1+δ2-δ1 is the deformation of the roadway roof during the two measurements. ④ There are two methods to measure the deformation of other parts of the top plate of this section. If the telescopic rod is of the ordinary type, the robot is moved to the area below the point to be measured. If it is of the rotating type, the robot is stationary, the telescopic rod is rotated and extended, and the oblique length and rotation angle are measured. Then, the actual vertical deformation is derived by using the relationship of trigonometric functions. ⑤ When measuring the floor heave of a roadway, the method in ③ can also be used, in which case the telescopic rod extends downwards; ⑥ The method for measuring the shrinkage of the two sides is the same as above, except that the telescopic rod extends to both sides.
6. The mining humanoid robot according to claim 5, characterized in that, After creating a 2D model of the tunnel cross-section, the existing fixed surrounding rock deformation monitoring system was used for verification. and / or To monitor the dynamic deformation of the surrounding rock, 3D modeling is performed based on the 2D modeling. The method is as follows: N measuring points are arranged in the tunnel, and 2D accurate modeling is performed on each measuring point one by one to construct a 3D model of the entire tunnel at time t0. Similarly, 3D modeling diagrams at times t1...tm are constructed and merged to form a dynamic 3D diagram of tunnel deformation with time as the axis.
7. The mining humanoid robot according to claim 1, characterized in that: The encryption measures used by the communication module are selected from at least one of the following sequences: low-strength encryption, medium-strength encryption, high-strength encryption, PQC hybrid mode, and pure PQC; and when an unauthorized access attempt to the communication channel is detected, the security controller can disable the motion control interface and only allow the transmission of a preset safety stop signal.
8. The mining humanoid robot according to claim 1, characterized in that, It also includes one or more of the following items: (1) The inspection range of each robot is determined by SLAM and the underground personnel positioning system, or by setting up signs at both ends of the roadway and at the bends of the branch roadway to indicate the activity range. The signs are signs, spray paint, or stickers, including one or a combination of line segments, characters, QR codes, barcodes, and radio frequency modules. (2) The combustible gas includes one or a combination of carbon monoxide, methane, gas, acetylene, natural gas, biogas, hydrogen sulfide, olefins, hydrogen, alcohol vapor, and ammonia. (3) The leg force safety threshold is 30% or more higher than the standard for normal walking leg lifting traction. (4) The robot body is equipped with a lifting lug for drone lifting, or is equipped with a lifting cage or ring sling.
9. The mining humanoid robot according to claim 1, characterized in that, It also includes at least one of the following: (1) The mobile module also includes a combination of switching between legged, wheeled, and tracked drive units; (2) The sensor data in the environmental perception module is provided to the core processing unit through the wireless communication module; (3) The operation execution module also includes a fire hose interface, a fire water tank and a high-pressure fine water mist extinguishing device; (4) The robot is also equipped with a fire extinguisher. The fire extinguisher nozzle is held by a universal clamping mechanism. The universal clamping mechanism is a multi-degree-of-freedom arm or a special clamping mechanism with no less than 2 degrees of freedom. The special clamping mechanism includes a quick-release buckle for clamping the nozzle, and X-axis and Y-axis universal clamping motors. (5) The tools used by the robot are mounted in the following ways: under normal conditions, one tool is connected to the robot's forearm and the other tools are placed in the toolbox; or a multi-tool turret is configured, with multiple tools radially connected to the turret, which is rotated to the target tool position when in use; (6) The robot’s special design to adapt to harsh working environments includes: a multi-seal structure for the charging port; anti-slip texture on the soles of the feet; impact-resistant buffer armor on the head, torso and joints; and built-in high-intensity light assembly. (7) The protective module further includes coating the outer shell with a heat-resistant coating and / or laying a heat insulation layer; (8) The front end of the air filter of the internal combustion engine or fuel cell is equipped with a cooling pipe network to cool the hot air drawn in. The cooling water is fire-fighting water or circulating cooling water.
