A multi-robot collaborative residual coal cleaning system and method based on visual guidance

Abstract

The present application relates to the technical field of cleaning the inner wall of a coal train gondola after unloading, and discloses a multi-robot arm collaborative residual coal cleaning system and method based on visual guidance, which comprises a ground rail assembly, a ground rail moving platform, a robot arm, a high-pressure supply device, an angle detection device, a laser radar detection device and a control system, and is based on laser radar three-dimensional point cloud data to identify the position of the support rod inside the gondola and divide it into four independent areas. When the overturning angle reaches a preset value, four robot arms are controlled to extend into their respective bound areas simultaneously, a dedicated blowing path is generated according to the distribution characteristics of the residual coal for high-pressure air blowing, and the working spaces of adjacent robot arms are constrained to be non-overlapping at the support rod boundary. The problems of low efficiency, high risk, high labor intensity of manual cleaning, easy damage to the gondola during mechanical cleaning, incomplete cleaning, energy waste and insufficient cooperation are solved, and the automation degree and operation safety of the residual coal cleaning of the gondola of the car dumper are improved.

Description

Technical Field

[0001] This invention relates to the field of cleaning technology for the inner walls of coal wagons after unloading coal, and particularly to a vision-guided multi-robotic arm collaborative residual coal cleaning system and method. Background Technology

[0002] In my country, bulk coal transportation mainly relies on railway open wagons, with the C80 type being a specialized transport vehicle for coal mines. Tippers are core equipment for unloading bulk coal in industries such as thermal power generation, ports, steel smelting, and mining. During the tipper unloading process, due to the inherent stickiness of coal and the roughness of the wagon's inner walls, coupled with the trapping effect of the tipper's pressing mechanism, a certain amount of residual coal remains on the inner walls, floor, around the support rods, and in various corners of the wagon after unloading. This results in a waste of coal resources, increased transportation energy consumption, and affects the quality of subsequent loading. Therefore, cleaning residual coal from the tipper wagon is an indispensable and crucial step in the coal unloading process.

[0003] Currently, the cleaning of residual coal in tippler wagons mainly relies on manual labor, which suffers from low efficiency, high risk, high labor intensity, rising costs, and difficulty in recruiting workers. With the development of automation technology and the industry's demand for safe production, cost reduction, and efficiency improvement, there is an urgent need for a residual coal cleaning system that can adapt to the C80 type open wagon structure, achieve multi-equipment collaboration, automatic adaptation, and efficient cleaning.

[0004] In existing technologies, mechanical roller brush or scraper cleaning methods employ contact cleaning, which easily damages the coating on the inner wall of the car body. Furthermore, the roller brushes wear out quickly, resulting in high maintenance costs. In rapping methods for cleaning residual coal, excessive rapping force can damage the car body structure, and the cleaning is incomplete, leaving corner coal difficult to remove. In high-pressure water jet methods, water mixes with coal to form a slurry, increasing the coal's moisture content and affecting its quality. Negative pressure suction methods require regular replacement of dust screens and are ineffective at removing highly adhesive residual coal. Installing robotic arms on the tipper beam increases the tipper's load, potentially causing structural deformation. In multi-robotic arm methods without designated zones, the lack of workspace division based on the car body's structural characteristics increases the risk of collisions or blind spots. Methods without visual perception cannot detect the distribution of residual coal in real time, hindering on-demand purging and resulting in energy waste. These existing technologies are insufficient in terms of cleaning effectiveness, equipment protection, operational coordination, and energy utilization, failing to meet the actual needs for efficient and automated cleaning of residual coal in tipper cars.

[0005] Therefore, there is an urgent need to provide a technical solution to address the above problems. Summary of the Invention

[0006] To address the aforementioned technical problems, this invention provides a vision-guided multi-robotic arm collaborative coal cleaning system and method.

[0007] Firstly, the present invention provides a vision-guided multi-robotic arm collaborative residual coal cleaning system, the technical solution of which is as follows: The ground rail assembly includes four sets of ground rails arranged in parallel within the coal bunker below the tipper, with the installation direction of the four sets of ground rails perpendicular to the length direction of the open wagon. Four ground-rail mobile platforms are respectively installed on four sets of ground rails and move in a straight line along the ground rails; Four robotic arms are fixed to the four ground-rail moving platforms respectively, and each robotic arm is equipped with a high-pressure nozzle at its end. A high-pressure supply device, connected to the high-pressure nozzle, is used to supply high-pressure air to the high-pressure nozzle; An angle detection device is installed on the rotating shaft of the tippler to detect the tipping angle of the tippler. A lidar detection device is used to scan the inner wall of the open wagon to obtain three-dimensional point cloud data of the inner wall of the wagon. The control system is electrically connected to the ground-rail moving platform, the robotic arm, the high-voltage supply device, the angle detection device, and the lidar detection device; the control system is used for: Based on the three-dimensional point cloud data, the positions of the three sets of transverse struts inside the open wagon are identified, and the interior of the wagon is divided into four independent areas according to the positions of the three sets of struts. The four robotic arms are then bound to the four independent areas one by one. When the tipping angle reaches the preset trigger angle, the four ground rail moving platforms are controlled to drive the four robotic arms to extend synchronously from the initial position of the side wall of the coal bunker along the ground rail into the area below their respective independent areas. Based on the three-dimensional point cloud data, the residual coal distribution characteristics of each independent region are extracted, and a dedicated purging path matching the residual coal distribution characteristics of the corresponding independent region is generated for each robotic arm. The dedicated purging path is selected from at least one of the following: serpentine scanning path, fixed-point focusing path, and edge-wall-attaching path. The four robotic arms are controlled to perform high-pressure air purging along their respective dedicated purging paths, and the end-effector working spaces of adjacent robotic arms are constrained to not overlap at the boundaries defined by the three sets of support rods.

[0008] The beneficial effects of the vision-guided multi-robotic arm collaborative residual coal cleaning system of the present invention are as follows: The system of this invention constructs a multi-robotic arm collaborative residual coal cleaning system comprising a ground rail assembly, a ground rail moving platform, robotic arms, a high-pressure supply device, an angle detection device, a lidar detection device, and a control system. Based on lidar three-dimensional point cloud data, it identifies the position of the support rods inside the car and divides it into four independent areas. When the tipping angle reaches a preset value, it controls four robotic arms to simultaneously extend into their respective bound areas. Combining the residual coal distribution characteristics, it generates a dedicated purging path for high-pressure air purging and constrains the working space of adjacent robotic arms to not overlap at the support rod boundaries. This solves the problems of low efficiency, high risk, and high labor intensity of existing manual cleaning, as well as the problems of easy damage to the car body, incomplete cleaning, energy waste, and insufficient coordination of mechanical cleaning. It improves the automation level and operational safety of residual coal cleaning in tipper cars.

[0009] Based on the above solution, the vision-guided multi-robotic arm collaborative coal cleaning system of the present invention can be further improved as follows.

[0010] In one alternative approach, the serpentine scanning path involves the high-pressure nozzle at the end of the robotic arm reciprocating along the width of an independent region and gradually advancing along its length. The relationship between the scanning interval and the effective purging radius of the high-pressure nozzle satisfies... ,in For scanning spacing, To achieve an effective purging radius, This is the overlap coefficient.

[0011] The advantages of adopting the above-mentioned optional method are as follows: by further setting the overlap coefficient relationship between the scanning interval and the effective purging radius, the high-pressure nozzle forms a reasonable purging coverage overlap during the reciprocating scanning process, avoiding the energy waste caused by purging blind spots and repeated purging, and improving the uniformity of residual coal cleaning and the utilization rate of high-pressure air under the serpentine scanning path.

[0012] In one optional approach, the fixed-point focusing path involves the high-pressure nozzle at the end of the robotic arm moving to the coordinates of a point in a localized area where the buildup thickness exceeds a preset threshold, then pausing for a preset time to perform intensive purging. The relationship between the dwell time and the buildup thickness satisfies... ,in For the duration of stay, Minimum stay time, For time coefficient, The thickness of the deposit.

[0013] The beneficial effects of adopting the above-mentioned optional method are as follows: by further establishing a linear relationship between residence time and accumulation thickness, the high-pressure nozzle can automatically extend the enhanced purging time in local areas where the residual coal accumulation is thicker, thereby achieving targeted cleaning of stubborn residual coal, avoiding the problems of incomplete cleaning or over-purging caused by fixed purging time, and improving the cleaning accuracy and adaptability of the fixed-point focusing path.

[0014] In one optional embodiment, the edge-attaching path involves the high-pressure nozzle at the end of the robotic arm moving along the contour line of the side wall of the carriage or the edge line of the floor plate. A preset distance is maintained between the high-pressure nozzle and the wall surface, and a preset angle is maintained between the blowing direction and the normal to the wall surface. The relationship between the moving speed of the high-pressure nozzle and the residual coal adhesion strength level satisfies... ,in For movement speed, As the reference speed, The velocity attenuation coefficient, This refers to the adhesion strength rating.

[0015] The beneficial effects of adopting the above-mentioned optional method are as follows: by further establishing an exponential decay relationship between the moving speed and the adhesion strength level of the residual coal, the high-pressure nozzle automatically reduces the moving speed in the wall area with high adhesion strength, ensuring the full stripping of residual coal in the edge and corner areas, avoiding cleaning omissions caused by uniform speed movement, and improving the cleaning quality of residual coal on complex wall surfaces by the edge-adhesive path.

[0016] In an alternative embodiment, the control system is further configured to: When any of the robotic arms needs to extend across the strut boundary to an adjacent independent area for purging operations, a pause command is sent to the adjacent robotic arm to keep it in a retracted posture. After the robotic arm that crossed the boundary completes the operation and returns to its independent area, a resume command is sent to the adjacent robotic arm.

[0017] The advantages of adopting the above-mentioned optional method are as follows: by further setting the pause and resume command interaction logic when crossing the strut boundary, adjacent robotic arms can form an interlocked avoidance in the boundary area, avoiding the spatial interference and collision risks when multiple robotic arms work together, and ensuring the continuity and safety of the purging operation at the strut boundary.

[0018] In one alternative embodiment, the ground rail is a triangular, right-angled, upward-facing structure, suspended from the bottom of the coal bunker by a bracket, with a preset distance maintained between the ground rail and the bottom of the coal bunker.

[0019] The advantages of adopting the above-mentioned optional method are as follows: by adopting a triangular right-angle upward structure and suspending it on the bottom surface of the coal bunker, a preset distance is formed between the ground rail and the bottom surface of the coal bunker, which avoids the impact of coal accumulation on the ground rail moving platform and ensures the long-term stable operation of the ground rail component under harsh working conditions.

[0020] In one alternative embodiment, a wear-resistant scraper is fixed to the bottom of the ground rail moving platform. The wear-resistant scraper is in contact with the right-angled surface of the ground rail and is used to scrape off the coal material on the surface of the ground rail when the ground rail moving platform moves.

[0021] The advantages of adopting the above-mentioned optional method are as follows: by fixing a wear-resistant scraper at the bottom of the ground rail moving platform and fitting it with the right angle surface of the ground rail, the scattered coal on the surface of the ground rail is scraped off in real time during the movement, which avoids the accumulation of coal causing increased running resistance of the ground rail moving platform or track blockage, and improves the self-cleaning ability and operational reliability of the ground rail system.

[0022] In one alternative embodiment, the control system is specifically used for: The start-up sequence of the four robotic arms is determined according to the purging duration of their respective independent areas. The robotic arm with the longer purging duration starts first, and the four robotic arms reset synchronously after purging is completed.

[0023] The beneficial effects of adopting the above-mentioned optional method are as follows: by further determining the start-up sequence of the robotic arms according to the purging duration and resetting synchronously after the purging is completed, the operation cycle of each robotic arm is optimized in a coordinated manner, avoiding the efficiency loss caused by some robotic arms finishing waiting too early or the overall operation time being uneven, and improving the overall rhythm consistency of multi-robotic arm collaborative operation.

[0024] In one alternative, the lidar detection device is mounted below the tipper or on the ground rail moving platform to scan the interior wall of the wagon before the robotic arm extends into the wagon.

[0025] The advantages of adopting the above-mentioned optional method are as follows: by further installing the lidar detection device under the tipper or on the ground rail moving platform and completing the scan before the robotic arm extends, the acquisition of three-dimensional point cloud data and the robotic arm purging operation are decoupled in time, avoiding mutual interference when scanning and purging are carried out simultaneously, and ensuring the accuracy and real-time performance of visual guidance data.

[0026] Secondly, this invention provides a vision-guided multi-robotic arm collaborative residual coal cleaning method, employing the vision-guided multi-robotic arm collaborative residual coal cleaning system provided by this invention. The technical solution of this method is as follows: Based on the three-dimensional point cloud data, the positions of the three sets of transverse struts inside the open wagon are identified, and the interior of the wagon is divided into four independent areas according to the positions of the three sets of struts. The four robotic arms are then bound to the four independent areas one by one. When the tipping angle reaches the preset trigger angle, the four ground rail moving platforms are controlled to drive the four robotic arms to extend synchronously from the initial position of the side wall of the coal bunker along the ground rail into the area below their respective independent areas. Based on the three-dimensional point cloud data, the residual coal distribution characteristics of each independent region are extracted, and a dedicated purging path matching the residual coal distribution characteristics of the corresponding independent region is generated for each robotic arm. The dedicated purging path is selected from at least one of the following: serpentine scanning path, fixed-point focusing path, and edge-wall-attaching path. The four robotic arms are controlled to perform high-pressure air purging along their respective dedicated purging paths, and the end-effector working spaces of adjacent robotic arms are constrained to not overlap at the boundaries defined by the three sets of support rods.

[0027] The beneficial effects of the vision-guided multi-robotic arm collaborative residual coal cleaning method of the present invention are as follows: The method of this invention solves the problems of low efficiency, high risk, and high labor intensity of existing manual cleaning, as well as the problems of easy damage to the car body, incomplete cleaning, energy waste, and insufficient coordination of mechanical cleaning, and improves the automation level and operational safety of residual coal cleaning in tipper cars.

[0028] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0029] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 This is a schematic diagram of an embodiment of a vision-guided multi-robotic arm collaborative residual coal cleaning system according to the present invention; Figure 2 This is a schematic diagram of the overall structure; Figure 3 This is a schematic diagram of the ground rail assembly structure; Figure 4 Workflow diagram; Figure 5 A schematic diagram of the open wagon's three sets of struts, four-section structure, and robotic arm distribution; Figure 6 Flowchart for lidar point cloud processing; Figure 7 This is a flowchart illustrating an embodiment of a vision-guided multi-robotic arm collaborative residual coal cleaning method according to the present invention. Detailed Implementation

[0030] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0031] Figure 1 This diagram illustrates a structural schematic of an embodiment of a vision-guided multi-robotic arm collaborative coal cleaning system provided by the present invention. Figure 1 As shown, the vision-guided multi-robotic arm collaborative coal cleaning system includes: The ground rail assembly includes four sets of ground rails arranged in parallel within the coal bunker below the tipper, with the installation direction of the four sets of ground rails perpendicular to the length direction of the open wagon.

[0032] The ground rail assembly refers to an integrated structure consisting of four sets of ground rails, supports, a transmission structure, and a drag chain system. The four sets of ground rails are arranged in parallel inside the coal bunker below the tipper, with the installation direction perpendicular to the length of the open wagon. For example, after the tipper flips an open wagon to 170°, the four sets of ground rails in the ground rail assembly correspond to four independent areas inside the open wagon separated by struts, providing a support foundation for the four robotic arms to move back and forth along the width of the open wagon.

[0033] A tippler is an industrial device that uses a rotating mechanism to rotate open railway wagons around their length axis to unload bulk materials. For example, after an open wagon carrying coal enters the tippler station, the tippler rotates the wagon from an initial 0° position with the wagon opening facing upwards to a 180° position with the wagon opening facing downwards, completing the coal unloading. A ground rail is a straight track installed in the coal bunker below the tippler. It is a triangular, right-angled upward structure made of high-strength steel and suspended by supports. It is used to support and guide the ground rail moving platform to move in a predetermined direction. For example, each set of ground rails consists of two parallel rails, suspended above the coal bunker grate. The length direction of the rails is perpendicular to the length direction of the open wagon, and the track wheels of the ground rail moving platform roll along the top surface of the ground rails. An open wagon is a railway freight car with an open top, used for transporting coal. The interior of the wagon is equipped with multiple sets of transverse struts as a reinforcing structure. For example, a coal mine open wagon is about 10m long and 2.9m wide. It has three sets of transverse round steel struts located in the upper middle part of the wagon, dividing the interior into areas A, B, C and D.

[0034] Four ground-rail mobile platforms are respectively installed on four sets of ground rails and move in a straight line along the ground rails.

[0035] Among them, the ground rail moving platform refers to a moving platform installed on the ground rail, which is driven by a servo motor and transmitted through gear and rack meshing, and moves linearly back and forth along the ground rail. It is used to carry the robotic arm and provide positional movement capability when the robotic arm is working. For example, after receiving the movement command from the control system, the ground rail moving platform on the second set of ground rails moves from the initial position on the side wall of the coal bunker along the ground rail to the predetermined working position below the open wagon area B.

[0036] Four robotic arms are fixed to the four ground-rail moving platforms, and each robotic arm is equipped with a high-pressure nozzle at its end.

[0037] The robotic arm refers to an industrial robotic arm with multiple joints and degrees of freedom, fixedly installed on a ground-rail moving platform. Its end effector is equipped with a high-pressure nozzle, and it executes a predetermined trajectory under the command of a control system to clean the residual coal on the inner walls of an open wagon. For example, the robotic arm bound to area B uses a 6-DOF joint structure. Based on the position of the ground-rail moving platform and the generated dedicated cleaning path, it drives the high-pressure nozzle at its end to scan and clean the bottom plate and side walls of area B. The high-pressure nozzle is a fluid ejection device installed at the end of the robotic arm and connected to a high-pressure supply device via pipeline. It is used to spray high-pressure air at a set speed and direction onto the area where residual coal adheres to the inner walls of the open wagon. For example, in a serpentine scanning path, the high-pressure nozzle performs a reciprocating scanning motion along the width of area B, spraying high-pressure air at a pressure of 5 MPa to blow the evenly distributed residual coal on the bottom plate away from the inner walls of the wagon.

[0038] A high-pressure supply device, connected to the high-pressure nozzle, is used to supply high-pressure air to the high-pressure nozzle.

[0039] The high-pressure supply device refers to a device that provides high-pressure air to the high-pressure nozzles, including a high-pressure air compressor, a pressure tank, a filter, and control valves. The output working pressure is adjustable within a set range. For example, when a lidar detection device detects localized coal accumulation in area C with a thickness exceeding a preset threshold, the control system increases the working pressure of the high-pressure supply device from 6MPa to 10MPa, providing a higher purging pressure for focused, enhanced purging. High-pressure air refers to clean compressed air that has been compressed by a high-pressure air compressor, stored in a pressure tank, and purified by a filter before being sprayed from the high-pressure nozzles. This clean compressed air serves as a non-contact purging medium for stripping residual coal from the inner walls of open wagons. For example, after the high-pressure air is pressurized to 8MPa by the high-pressure supply device, it is sprayed at a set angle from the high-pressure nozzles at the end of a robotic arm onto the surface of residual coal adhering to the edge of the wagon's side wall. The residual coal detaches from the wall under the impact of the high-pressure airflow and falls into the coal bunker below.

[0040] An angle detection device is installed on the rotating shaft of the tipper and is used to detect the tipping angle of the tipper.

[0041] Among them, the angle detection device refers to a high-precision angle sensor installed on the rotating shaft of the tipper, which is used to detect the rotation angle of the tipper in real time and transmit the angle signal to the control system. For example, when the tipper drives the open car to start rotating and unloading coal from 0°, the angle detection device continuously sends the real-time tipping angle value to the control system. When the tipping angle reaches 170°, the control system determines that the tipping has been completed.

[0042] The rotating shaft refers to the mechanical shaft on the tipper used to drive the open wagon to rotate around its length. For example, the rotating shaft of the tipper is located at the bottom of both sides of the open wagon, and the angle detection device is fixed to the end of the rotating shaft, rotating synchronously with the rotating shaft and detecting the rotation angle. The tipping angle refers to the rotation angle of the tipper's rotating shaft relative to the horizontal plane. The initial state is 0° when the open wagon's opening is facing upwards. As the rotation angle of the tipper gradually increases, the open wagon gradually tilts to the side and eventually emptys. For example, at 0°, the open wagon's opening is facing upwards and is in a waiting state. When the tipping angle reaches 170°, the open wagon's opening is almost completely facing downwards, and the main coal in the wagon has been unloaded into the coal bunker by gravity, with only a small amount of residual coal adhering to the inner wall to be cleaned.

[0043] A lidar detection device is used to scan the inner wall of the open wagon to obtain three-dimensional point cloud data of the inner wall of the wagon.

[0044] Among them, the lidar detection device refers to a detection device that uses three-dimensional lidar to generate three-dimensional point cloud data of the inner wall of the open wagon by emitting a laser beam and receiving the reflected signal. This data is used to obtain information such as the position of the support rod, the position of the remaining coal, the thickness of the pile, and the level of adhesion strength. For example, after the tipper tipps the wagon into place, the lidar detection device completes a rapid scan of the inner wall of the open wagon within 5 seconds, generating a three-dimensional spatial point set that includes the side walls, bottom plate, base of the support rod, and corners of the wagon.

[0045] The inner wall of the wagon refers to all surfaces inside the wagon's interior, including the floor, side walls, front and rear end walls, and the base and corner areas of the struts. For example, after the open wagon is unloaded by the tipper, the central area of ​​the floor of the wagon's inner wall is mostly empty of coal, but residual coal of varying thicknesses remains at the edges where the side walls meet the floor, around the base of the struts, and in the corners. Three-dimensional point cloud data refers to the set of three-dimensional spatial points on the inner wall of the open wagon's interior obtained by a lidar detection device. Each point contains X, Y, and Z coordinates and reflection intensity information, used for subsequent strut identification, region segmentation, and residual coal distribution feature extraction. For example, the control system filters and removes outliers from the point cloud data consisting of approximately 100,000 three-dimensional spatial points obtained by lidar scanning, and then extracts the spatial coordinates of the struts and the height distribution of residual coal on the floor and side walls in each region.

[0046] The control system is electrically connected to the ground-rail moving platform, the robotic arm, the high-voltage supply device, the angle detection device, and the lidar detection device; the control system is used for: Based on the three-dimensional point cloud data, the positions of the three sets of transverse struts inside the open wagon are identified, and the interior of the wagon is divided into four independent areas according to the positions of the three sets of struts. The four robotic arms are then bound to the four independent areas one by one.

[0047] The positions of the three sets of transverse struts refer to the spatial coordinates of the three sets of transverse round steel struts inside the open wagon, identified by analyzing 3D point cloud data. These struts are arranged laterally along the width of the open wagon and are located in the upper middle part of the wagon compartment. For example, the control system will... The coordinates are identified as follows: , and Using the centerline plane of each strut as the dividing boundary, the open wagon body is sequentially divided along its length into areas A, B, C, and D. Each independent area refers to one of four working zones within the wagon body, separated by the centerline plane of the three sets of transverse struts. Each independent area is equipped with a robotic arm for residual coal cleaning. For example, area B is located between the first and second sets of struts, approximately 2.8m long and 2.9m wide. The second robotic arm is responsible for cleaning residual coal from the floor and side walls of area B. The end effector space of the second robotic arm is limited to area B. Within the coordinate range.

[0048] When the tipping angle reaches the preset trigger angle, the four ground rail moving platforms are controlled to drive the four robotic arms to extend synchronously from their initial positions on the side wall of the coal bunker into their respective independent areas along the ground rails.

[0049] The preset trigger angle refers to the tipping angle threshold that is pre-set and stored in the control system. When the tipping angle detected by the angle detection device reaches this threshold, the control system determines that the tipping has been completed and triggers the subsequent laser radar scanning and robotic arm operation process. For example, if the preset trigger angle is set to 165°, when the tipping machine flips the open wagon to 165°, the angle detection device sends a position signal to the control system, and the control system then starts the laser radar scanning program.

[0050] The "side wall" of the coal bunker refers to the vertical wall on the side of the coal bunker below the tipper. The robotic arms and ground-rail moving platforms are initially retracted and positioned at this location. For example, before unloading begins, all four ground-rail moving platforms and four robotic arms are stationary at their initial positions on the side wall of the coal bunker, with the joints of the robotic arms retracted and folded to clear the unloading passage for the open wagons. The initial position refers to the position of the ground-rail moving platforms and robotic arms when not in operation, located at the end of the ground rail closest to the side wall of the coal bunker. For example, after the tipper has reversed, the four ground-rail moving platforms move in the opposite direction along the ground rail back to their initial positions on the side wall of the coal bunker, with the robotic arms retracted and folded, awaiting the next open wagon to enter the tipping position.

[0051] Based on the three-dimensional point cloud data, the residual coal distribution characteristics of each independent region are extracted, and a dedicated purging path matching the residual coal distribution characteristics of the corresponding independent region is generated for each robotic arm. The dedicated purging path is selected from at least one of the following: serpentine scanning path, fixed-point focusing path, and edge-attached path.

[0052] Among them, the residual coal distribution characteristics refer to the information on the attachment location, coverage area, accumulation thickness and adhesion strength level of residual coal in each independent area extracted by analyzing the three-dimensional point cloud data, which is used to determine the type and parameters of the dedicated purging path; for example, after the lidar detection device acquires the three-dimensional point cloud data of area C, the control system extracts the residual coal distribution characteristics of area C as uniformly distributed on the bottom plate, covering an area of ​​about 70% of the area, with an average accumulation thickness of about 2cm and an adhesion strength level of loose, and generates a serpentine scanning path for area C accordingly.

[0053] Among them, the dedicated purging path refers to the specific purging motion trajectory generated by the control system for the corresponding robotic arm in a single independent area based on the distribution characteristics of the remaining coal in that area. It is selected from at least one of the following: serpentine scanning path, fixed-point focusing path, and edge-wall-attaching path. For example, if the remaining coal distribution characteristics of area A are that the bottom plate is uniformly distributed, the control system generates a serpentine scanning path for the No. 1 robotic arm in area A as the dedicated purging path. The high-pressure nozzle at the end of the No. 1 robotic arm reciprocates along the width direction of area A and gradually advances along the length direction.

[0054] The serpentine scanning path refers to a purging path in which the high-pressure nozzle at the end of the robotic arm performs a reciprocating scanning motion along the width of an independent area and gradually advances along its length. This is suitable for areas where residual coal is evenly distributed in a planar pattern. For example, if the residual coal coverage area in area A is approximately 65% ​​of the area and has a uniform thickness, a serpentine scanning path is used for the dedicated purging path. The nozzle reciprocates along the width of area A in a back-and-forth manner, advancing the scanning distance along the length after each back-and-forth motion. , according to Calculation, where The overlap coefficient is set to 0.7. The effective purging radius is defined as follows: A fixed-point focusing path refers to the path where the high-pressure nozzle at the end of the robotic arm moves to a local point where the coal accumulation thickness exceeds a preset threshold and then remains there for a preset time for intensive purging. This is suitable for areas with severe coal accumulation. For example, in area C, there is a local area on the floor with a coal accumulation thickness of 8cm, exceeding the preset threshold by 5cm. A fixed-point focusing path is used for this purpose, with the nozzle remaining at that point for a specified time. according to Calculation, where The minimum dwell time is set to 2 seconds. Take the time coefficient , Assuming a stacking thickness of 8cm, we get Intensive purging is performed. The edge-attached path refers to the path along which the high-pressure nozzle at the end of the robotic arm moves along the contour line of the side wall or the edge line of the floor of the carriage. A preset distance is maintained between the nozzle and the wall, and the purging direction maintains a preset angle with the wall normal. This is suitable for areas where residual coal adheres to the side wall or edge. For example, in area B, residual coal adheres to the edge where the left side wall meets the floor, with a medium adhesion strength level. The dedicated purging path uses the edge-attached path, maintaining an 80mm distance between the nozzle and the side wall, a 30° angle between the purging direction and the wall normal, and a nozzle movement speed... according to Confirmed, among which As the reference speed, The velocity attenuation coefficient, Assuming an adhesion strength rating of 2, the calculated moving speed is reduced to approximately 60% of the baseline speed for wall-mounted purging.

[0055] The four robotic arms are controlled to perform high-pressure air purging along their respective dedicated purging paths, and the end-effector working spaces of adjacent robotic arms are constrained to not overlap at the boundaries defined by the three sets of support rods.

[0056] The end-effector's working space refers to the spatial range that the high-pressure nozzle at the end of each robotic arm can reach during operation, which, under the constraints of the control system, does not exceed the boundary of the bound independent area; for example, the end-effector's working space is restricted to the area corresponding to region B. Within the coordinate range, the end effector space of the third robotic arm is limited to the region corresponding to C. Within the coordinate range, the end-effector working spaces of the two robotic arms do not overlap at the centerline plane of the strut between region B and region C.

[0057] The technical solution of this embodiment constructs a multi-robotic arm collaborative residual coal cleaning system, which includes a ground rail assembly, a ground rail moving platform, robotic arms, a high-pressure supply device, an angle detection device, a lidar detection device, and a control system. Based on lidar three-dimensional point cloud data, it identifies the position of the support rods inside the car and divides it into four independent areas. When the tipping angle reaches a preset value, it controls four robotic arms to simultaneously extend into their respective bound areas. Combined with the residual coal distribution characteristics, it generates a dedicated blowing path for high-pressure air blowing and constrains the working space of adjacent robotic arms to not overlap at the support rod boundaries. This solves the problems of low efficiency, high risk, and high labor intensity of existing manual cleaning, as well as the problems of easy damage to the car, incomplete cleaning, energy waste, and insufficient coordination of mechanical cleaning. It improves the automation level and operational safety of residual coal cleaning in tipper cars.

[0058] In one alternative approach, the serpentine scanning path involves the high-pressure nozzle at the end of the robotic arm reciprocating along the width of an independent region and gradually advancing along its length. The relationship between the scanning interval and the effective purging radius of the high-pressure nozzle satisfies... ,in For scanning spacing, To achieve an effective purging radius, This is the overlap coefficient.

[0059] In the above-mentioned optional methods, by further setting the overlap coefficient relationship between the scanning interval and the effective purging radius, the high-pressure nozzle forms a reasonable purging coverage overlap during the reciprocating scanning process, avoiding purging blind spots and energy waste caused by repeated purging, and improving the uniformity of residual coal cleaning and the utilization rate of high-pressure air under the serpentine scanning path.

[0060] In one optional approach, the fixed-point focusing path involves the high-pressure nozzle at the end of the robotic arm moving to the coordinates of a point in a localized area where the buildup thickness exceeds a preset threshold, then pausing for a preset time to perform intensive purging. The relationship between the dwell time and the buildup thickness satisfies... ,in For the duration of stay, Minimum stay time, For time coefficient, The thickness of the deposit.

[0061] In the above-mentioned optional methods, by establishing a linear relationship between residence time and accumulation thickness, the high-pressure nozzle can automatically extend the enhanced purging time in local areas where the residual coal accumulation is thicker, thereby achieving targeted cleaning of stubborn residual coal and avoiding the problems of incomplete cleaning or over-purging caused by fixed purging time, thus improving the cleaning accuracy and adaptability of the fixed-point focusing path.

[0062] In one optional embodiment, the edge-attaching path involves the high-pressure nozzle at the end of the robotic arm moving along the contour line of the side wall of the carriage or the edge line of the floor plate. A preset distance is maintained between the high-pressure nozzle and the wall surface, and a preset angle is maintained between the blowing direction and the normal to the wall surface. The relationship between the moving speed of the high-pressure nozzle and the residual coal adhesion strength level satisfies... ,in For movement speed, As the reference speed, The velocity attenuation coefficient, This refers to the adhesion strength rating.

[0063] In the above-mentioned optional methods, by further constructing an exponential decay relationship between the moving speed and the adhesion strength level of the residual coal, the high-pressure nozzle automatically reduces the moving speed in the wall area with high adhesion strength, ensuring the full stripping of residual coal in the edge and corner areas, avoiding cleaning omissions caused by uniform speed movement, and improving the cleaning quality of residual coal on complex wall surfaces by the edge-adhesive path.

[0064] In an alternative embodiment, the control system is further configured to: When any of the robotic arms needs to extend across the strut boundary to an adjacent independent area for purging operations, a pause command is sent to the adjacent robotic arm to keep it in a retracted posture. After the robotic arm that crossed the boundary completes the operation and returns to its independent area, a resume command is sent to the adjacent robotic arm.

[0065] The pause command refers to the instruction sent by the control system to the robotic arm to pause its current movement and maintain its current position. This is used to achieve interlocking avoidance when crossing boundaries in multi-robotic arm collaborative operations. For example, when robotic arm 1 needs to briefly cross the boundary to enter area B while cleaning the dead corner at the base of the first set of support rods, the control system sends a pause command to robotic arm 2, which is responsible for area B. Upon receiving the pause command, robotic arm 2 immediately stops its current purging action and maintains its current position. The retracted posture refers to the robotic arm retracting its joints to a folded state close to its body after receiving a pause command or completing the purging operation, thereby reducing the coverage area of ​​the end effector. For example, after receiving a pause command, robotic arm 2's six joints simultaneously retract to a folded state, and the end effector high-pressure nozzle retracts to a position close to the robotic arm base, providing a safe space for robotic arm 1 to cross the support rod boundary and enter area B for purging operations.

[0066] The recovery command refers to the command sent by the control system to the robotic arm that previously received the pause command to restart the operation. This command is issued after the robotic arm that is working across the boundary completes the operation and returns to the bounding area. For example, after the first robotic arm completes the cleaning operation at the root of the first set of support rods and returns to area A, the control system sends a recovery command to the second robotic arm. The second robotic arm then unfolds from the retracted posture and continues to perform the previously interrupted fixed-point focusing purging operation.

[0067] In the above-mentioned optional methods, by further setting the pause and resume command interaction logic when crossing the strut boundary, adjacent robotic arms can form an interlocked avoidance in the boundary area, avoiding the spatial interference and collision risks when multiple robotic arms work together, and ensuring the continuity and safety of the purging operation at the strut boundary.

[0068] In one alternative embodiment, the ground rail is a triangular, right-angled, upward-facing structure, suspended from the bottom of the coal bunker by a bracket, with a preset distance maintained between the ground rail and the bottom of the coal bunker.

[0069] The triangular right-angled upward structure refers to a structure where the cross-section of the ground rail is triangular with the right-angled sides pointing upwards. The ground rail is suspended by supports, and its top surface forms a horizontal bearing surface for the track wheels of the ground rail moving platform to travel on. The inclined surface allows scattered coal to slide off naturally. For example, with the right-angled sides of the ground rail pointing vertically upwards, the track wheels roll on the horizontal surface of the ground rail's top surface. During coal unloading, coal scattered on the inclined surface of the ground rail automatically slides off to the bottom of the coal bunker under gravity, preventing coal from accumulating on the ground rail surface. The preset spacing refers to the predetermined vertical distance maintained between the bottom surface of the ground rail and the bottom surface of the coal bunker after the ground rail is suspended by supports, used to prevent coal from burying the ground rail. For example, the bottom surface of the triangular structure of the ground rail is supported by supports with a support height set at 12cm, forming a 12cm suspended space between the bottom surface of the ground rail and the surface of the coal bunker grate. Scattered coal slides off the triangular inclined surface and falls directly from the suspended space into the coal conveying channel at the bottom of the coal bunker.

[0070] In the above-mentioned optional methods, a triangular right-angle upward structure is further adopted and suspended on the bottom surface of the coal bunker, so that a preset distance is formed between the ground rail and the bottom surface of the coal bunker. This avoids the impact of coal accumulation on the ground rail moving platform and ensures the long-term stable operation of the ground rail component under harsh working conditions.

[0071] In one alternative embodiment, a wear-resistant scraper is fixed to the bottom of the ground rail moving platform. The wear-resistant scraper is in contact with the right-angled surface of the ground rail and is used to scrape off the coal material on the surface of the ground rail when the ground rail moving platform moves.

[0072] Among them, the wear-resistant scraper refers to a plate-shaped scraper fixed to the bottom of the ground rail moving platform. It is made of wear-resistant material, and the edge of the scraper is in contact with the right-angle surface of the ground rail. During the movement of the ground rail moving platform, it scrapes off the coal material accumulated on the surface of the ground rail. For example, the wear-resistant scraper at the bottom of the ground rail moving platform is made of polyurethane material with a thickness of 10mm. The gap between the lower edge of the scraper and the vertical right-angle surface of the ground rail is no more than 1mm. When the ground rail moving platform moves from the initial position to the working position, the wear-resistant scraper scrapes off the small amount of coal material remaining above the right-angle surface of the ground rail. The scraped coal material falls into the bottom of the coal bunker from the suspended space.

[0073] In the above-mentioned optional methods, by further fixing a wear-resistant scraper to the bottom of the ground rail moving platform and fitting it into the right-angled surface of the ground rail, the scattered coal material on the surface of the ground rail is scraped off in real time during the movement, which avoids the accumulation of coal material leading to increased running resistance of the ground rail moving platform or track blockage, and improves the self-cleaning ability and operational reliability of the ground rail system.

[0074] In one alternative embodiment, the control system is specifically used for: The start-up sequence of the four robotic arms is determined according to the purging duration of their respective independent areas. The robotic arm with the longer purging duration starts first, and the four robotic arms reset synchronously after purging is completed.

[0075] The purging time refers to the total time required for a single robotic arm to complete the cleaning of residual coal in a bound independent area. It is determined by the length of the dedicated purging path, the nozzle movement speed, and the dwell time, which are determined by the distribution characteristics of residual coal in the corresponding area. For example, the residual coal distribution characteristics of area D are that there is local accumulation on the bottom plate and residual coal attached to the side wall edge. The dedicated purging path is a combination of a fixed-point focusing path and an edge-attached path. The purging time is calculated to be 25s. The purging time of area A is 18s, the purging time of area B is 22s, and the purging time of area C is 30s.

[0076] Synchronous reset refers to the synchronized action of all four robotic arms retracting from their respective end positions and returning to their initial positions on the side wall of the coal bunker after completing their respective purging operations. For example, after the third robotic arm, which has the longest purging time, completes 30 seconds of purging, the control system confirms that all four robotic arms have completed their operations and then issues a synchronous reset command. The four robotic arms simultaneously retract their joints and return to their initial positions on the side wall of the coal bunker along the ground rail.

[0077] In the above-mentioned optional methods, the starting sequence of the robotic arms is further determined by the purging duration and the robotic arms are reset synchronously after purging is completed. This optimizes the operation cycle of each robotic arm, avoids the efficiency loss caused by some robotic arms finishing waiting too early or the overall operation time being uneven, and improves the overall rhythm consistency of multi-robotic arm collaborative operation.

[0078] In one alternative, the lidar detection device is mounted below the tipper or on the ground rail moving platform to scan the interior wall of the wagon before the robotic arm extends into the wagon.

[0079] In the above-mentioned optional methods, by further installing the lidar detection device under the tipper or on the ground rail moving platform and completing the scan before the robotic arm extends, the acquisition of 3D point cloud data and the robotic arm purging operation are decoupled in time, avoiding mutual interference when scanning and purging are carried out simultaneously, and ensuring the accuracy and real-time nature of the visual guidance data.

[0080] like Figure 2 As shown, the ground rail assembly is located inside the coal bunker below the tipper. The assembly comprises four sets of parallel ground rails, installed perpendicular to the length of the open wagon. Each set consists of two tracks made of high-strength steel with a triangular cross-section and the right-angled sides pointing upwards. These tracks are suspended from the bottom of the coal bunker by supports, maintaining a pre-set distance between them to prevent coal from burying them. The ground rails are equipped with a rack and pinion drive structure and a bridge-type fully enclosed cable chain system. This system houses the power cables, signal lines, and high-voltage lines of the robotic arm.

[0081] like Figure 3 As shown, the structure of the ground rail assembly includes a ground rail moving platform, a robotic arm base, track wheels, and a drag chain system. Four ground rail moving platforms are respectively installed on four sets of ground rails. The moving platforms are driven by servo motors and transmit power through gear and rack meshing, moving linearly back and forth along the ground rails via the track wheels. Each ground rail moving platform has a wear-resistant scraper fixed to its bottom. The scraper material can be polyurethane, and it is fitted to the right-angled surface of the ground rail with a gap of no more than 1mm. This scraper is used to simultaneously scrape away coal accumulated on the surface of the ground rail during the platform's movement. The scraped coal falls directly from the suspended position to the coal conveying channel at the bottom of the coal bunker. Limit switches are installed on the ground rail moving platforms to ensure a smooth stop when the platform reaches its designated position.

[0082] Four robotic arms are vertically fixed to four ground-rail moving platforms. Each robotic arm adopts a 6-DOF industrial robotic arm structure and is dustproof and waterproof. A high-pressure nozzle is fixedly installed at the end of each robotic arm, and the high-pressure nozzle is connected to the high-pressure supply device through pipelines. The four robotic arms correspond one-to-one with the four independent areas inside the open wagon, which are divided by three sets of transverse struts. Each robotic arm is responsible for cleaning the remaining coal in one independent area.

[0083] The high-pressure supply device includes a high-pressure air compressor, a pressure tank, a filter, pipelines, and control valves. Flow sensors and pressure sensors are installed on the pipelines. The high-pressure supply device provides high-pressure air to the high-pressure nozzles at the ends of the four robotic arms. The high-pressure air is used as a non-contact purging medium, and the working pressure is adjustable in the range of 0.6MPa to 15MPa.

[0084] The angle detection device uses a high-precision angle sensor, which is installed on the rotating shaft of the tipper. The angle detection device detects the tipping angle in real time and transmits the angle signal to the control system.

[0085] The lidar detection device employs a 3D lidar system, installed below the tipper or on a ground-rail moving platform. The lidar system generates 3D point cloud data of the inner wall of the open wagon by emitting a laser beam and receiving reflected signals, scanning the inner wall before the robotic arm extends into the wagon. The processed 3D point cloud data is used to extract information such as strut position coordinates, residual coal attachment location, accumulation thickness, and adhesion strength level. Specifically, the strut position coordinates are used to identify the spatial positions of the three sets of struts and serve as the boundary for partitioning. The residual coal attachment location is used to identify the distribution area of ​​residual coal on the side walls, floor, strut roots, and corners of the wagon. The accumulation thickness is calculated from the point cloud height difference, and the adhesion strength level is determined based on the point cloud reflection intensity or historical data models to assess the degree of adhesion of the residual coal.

[0086] The control system uses a PLC or DCS controller as the core control unit, equipped with a touch screen interface, enabling parameter setting, status monitoring, fault alarms, and manual / automatic switching. The control system incorporates a spatial modeling unit, a path generation unit, and a collaborative control unit. It is electrically connected to the angle detection device, the servo motors of the four ground-rail moving platforms, the drive units of the four robotic arms, the control valves of the high-pressure supply device, and the lidar detection device.

[0087] like Figure 4 As shown, the workflow includes the initial state, tipper tipping, tipper tipping into position, angle sensor sending tipping into position signal, lidar scanning of the inner wall of the car, control system extracting the position of the support rod and dividing it into four independent areas, extracting the position, thickness and adhesion level of residual coal in each zone, generating the blowing path of each robotic arm, ground rail moving platform driving robotic arm to perform coordinated blowing operation below the corresponding zone, high pressure air shut off after blowing is completed, robotic arm returns to the initial state, moving platform returns to the initial position, tipper tipping back and completing this tipping and unloading.

[0088] Specifically, the control system implements spatial modeling and partition binding based on LiDAR according to the following process: After the tipper reaches its designated position and before the robotic arm extends, a lidar detection device performs a 3D scan of the inner wall of the open wagon's cargo compartment, acquiring point cloud data of the interior. The control system filters, segments, and extracts features from the point cloud data, identifying the spatial coordinates of three sets of struts, which appear as transverse protrusions along the width of the wagon. Using the centerline plane of each strut as a natural dividing boundary, the control system divides the interior of the wagon into four independent regions, denoted as Region A, Region B, Region C, and Region D. The length of each independent region is determined by the distance between adjacent struts.

[0089] like Figure 5 As shown, the open wagon's interior has three sets of transverse struts. The area between struts 1 and 2 is region A; the area between struts 2 and 3 is regions B and C; and the area between strut 3 and the end of the wagon is region D. A bathtub structure is located at the bottom of the wagon. Robotic arm 1 is attached to region A, robotic arm 2 to region B, robotic arm 3 to region C, and robotic arm 4 to region D. The end-effector's working space is strictly limited to the boundaries of its corresponding independent region. Its X-direction movement is restricted to the area between the starting and ending boundary coordinates of its corresponding independent region and must not cross the centerline plane of adjacent struts.

[0090] Furthermore, the control system processes the point cloud data within each independent area to extract the location, thickness, and strength of residual coal attachment. The location of residual coal attachment includes the sidewalls, bottom plate, strut roots, and corners. The thickness is calculated using the difference between the point cloud height and the empty car reference height. The strength of attachment is classified as loose, medium, or compacted using point cloud reflection intensity analysis or a machine learning model.

[0091] like Figure 6 As shown, the LiDAR point cloud processing flow is as follows: First, LiDAR point cloud acquisition is performed, followed by point cloud preprocessing, which includes filtering, outlier removal, and coordinate system transformation. Then, point cloud segmentation and feature extraction are performed to calculate the strut position coordinates. Next, the partition boundaries are calculated, dividing the area into four independent regions. Then, the point clouds of the base plate and side walls are differentiated from the empty vehicle reference point cloud to calculate the stacking thickness. Finally, point cloud reflection intensity analysis is performed to calculate the adhesion strength level.

[0092] Based on the distribution characteristics of residual coal in each zone detected by lidar, the control system automatically selects or combines three basic path modes—serpentine scanning path, fixed-point focusing path, and edge-wall-attaching path—to generate a unique purging path for each robotic arm.

[0093] The serpentine scanning path is suitable for scenarios where the remaining coal within a zone is evenly distributed in a planar pattern, meaning the remaining coal coverage area exceeds 60% and the thickness is uniform. The serpentine scanning path involves the high-pressure nozzle at the end of the robotic arm performing a reciprocating scanning motion along the width of an independent zone and gradually advancing along its length, forming a zigzag coverage trajectory. Scanning interval... Effective purging radius of high-pressure nozzle Between satisfy ,in This is the overlap coefficient, ranging from 0.6 to 0.8, ensuring 20% ​​to 40% overlap between adjacent trajectories to eliminate purge blind spots. Overlap coefficient The control system automatically adjusts the value based on the adhesion strength of the remaining coal, using a smaller value when the adhesion strength is high and a larger value when the adhesion strength is low. Effective purging radius of the high-pressure nozzle. It is determined by the nozzle type and the working air pressure.

[0094] The fixed-point focusing path is suitable for scenarios where there is severe coal accumulation in localized areas within a zone, specifically when the lidar detects that the accumulation thickness exceeds a preset threshold. The fixed-point focusing path involves the high-pressure nozzle at the end of the robotic arm moving to the coordinates of the point where the accumulation thickness exceeds the preset threshold, pausing for a preset time for intensive purging, and then moving to the next key area. Key areas are processed sequentially from thickest to thinnest accumulation. (Dwell time...) With stacking thickness Between satisfy ,in Minimum stay time, This is a time coefficient. Minimum stay time. and time coefficient The stacking thickness was determined experimentally. It is calculated from the height difference of the point cloud by lidar.

[0095] Edge-attached wall path is suitable for scenarios where residual coal adheres to the side wall or floor edge of the carriage, i.e., when the lidar detects a high degree of overlap between the residual coal point cloud and the wall or edge. The edge-attached wall path involves the high-pressure nozzle at the end of the robotic arm moving along the side wall contour or floor edge line, maintaining a preset distance between the high-pressure nozzle and the wall, and maintaining a preset angle between the blowing direction and the wall normal to ensure tangential airflow impact on the wall. High-pressure nozzle movement speed... Adhesion strength grade of residual coal Between satisfy ,in As the reference speed, The velocity attenuation coefficient represents the adhesion strength grade. The larger the value, the slower the movement speed and the longer the purging time. Adhesion strength rating The reference speed is determined by the lidar reflection intensity or historical data model. and velocity decay coefficient Calibrated by experiments.

[0096] The automatic path selection logic is as follows: when the detected residual coal coverage area exceeds 60% and the thickness is uniform, a serpentine scanning path is selected; when there is localized accumulation exceeding a preset threshold, a fixed-point focusing path is selected; when the attachment location is a sidewall or edge, an edge-attaching path is selected; in mixed situations, the fixed-point focusing path is used first to clean the thick accumulation, and then the serpentine scanning path is used to cover the remaining area. When switching from the fixed-point focusing path to the serpentine scanning path, the high-pressure nozzle at the end of the robotic arm moves directly from the last fixed-point focusing coordinate to the starting point of the serpentine scanning path, and the high-pressure nozzle remains closed during the movement. The fixed-point focusing path and the edge-attaching path can be used alone or in combination with the serpentine scanning path.

[0097] The control system also implements multi-robotic arm collaborative control and interlocking mechanisms, including spatial isolation constraints, time-coordinated scheduling, and cross-boundary operation interlocking.

[0098] The spatial isolation constraint is based on partition boundary binding. The end effectors of the four robotic arms are always located within the X coordinate range of their respective independent regions, and a safe distance of at least one strut width is maintained between adjacent robotic arms, fundamentally eliminating the risk of collision.

[0099] The time-coordinated scheduling sets the start priority based on the purging time of each independent area. The purging time is calculated by comprehensively considering the remaining coal amount, accumulation thickness and adhesion strength of the corresponding independent area. The robotic arm with a longer purging time starts first, and the robotic arm with a shorter purging time starts later. The start time difference between adjacent robotic arms ranges from 0.5s to 2s. After the purging is completed, the four robotic arms retract and reset synchronously.

[0100] The cross-boundary operation interlock is implemented when any robotic arm needs to extend beyond the strut boundary to an adjacent independent area for purging operations, such as cleaning a dead corner at the base of the strut that crosses the partition boundary. The control system sends a pause command to the adjacent robotic arm, which remains in a retracted position. Once the robotic arm that crossed the boundary completes its operation and returns to its designated independent area, the control system sends a resume command to the adjacent robotic arm. During the interlock, the movement speed of the robotic arm crossing the boundary is reduced to half its normal speed. The reduction ratio is determined based on the relationship between the time required for the adjacent robotic arm to retract and the time it takes for the robotic arm crossing the boundary to reach the strut centerline plane, ensuring that the adjacent robotic arm completes its retraction switch before the robotic arm crossing the boundary reaches the strut centerline plane.

[0101] The control system also coordinates with the tipping process. An angle detection device monitors the tipping angle of the tipper in real time. When the tipping angle reaches the preset trigger angle, the control system determines that the tipping is in place and the main coal in the car has been unloaded, with the remaining coal awaiting cleaning. It then activates the lidar detection device for a rapid scan, taking less than 5 seconds, to obtain the position of the support rods and the distribution characteristics of the remaining coal. Afterward, the control system activates four ground-rail moving platforms, synchronously moving four robotic arms from their initial positions on the side walls of the coal bunker to their respective working positions below their respective independent areas. After purging is completed, the control system sends a return tipping command to the tipper, which returns to its initial position, ready for the next open car operation.

[0102] Figure 7 This diagram illustrates a flow chart of an embodiment of a vision-guided multi-robotic arm collaborative residual coal cleaning method provided by the present invention. The method employs a vision-guided multi-robotic arm collaborative residual coal cleaning system as provided by the present invention. Figure 7 As shown, the method includes the following steps: S1. Based on the three-dimensional point cloud data, identify the positions of the three sets of transverse struts inside the open wagon, and divide the interior of the wagon into four independent areas according to the positions of the three sets of struts, and bind the four robotic arms to the four independent areas one by one; S2. When the tipping angle reaches the preset trigger angle, control the four ground rail moving platforms to drive the four robotic arms to extend synchronously from the initial position of the side wall of the coal bunker along the ground rail into the area below their respective independent areas. S3. Based on the three-dimensional point cloud data, extract the residual coal distribution characteristics of each independent region, and generate a dedicated purging path for each robotic arm that matches the residual coal distribution characteristics of the corresponding independent region. The dedicated purging path is selected from at least one of the following: serpentine scanning path, fixed-point focusing path, and edge-attached path. S4. Control the four robotic arms to perform high-pressure air purging along their respective dedicated purging paths, and constrain the end-effector working spaces of adjacent robotic arms to not overlap at the boundaries defined by the three sets of support rods.

[0103] It should be noted that the beneficial effects of the vision-guided multi-robotic arm collaborative coal cleaning method provided in the above embodiments are the same as those of the vision-guided multi-robotic arm collaborative coal cleaning system described above, and will not be repeated here. Furthermore, the system provided in the above embodiments is only illustrated by the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the system can be divided into different functional modules according to the actual situation to complete all or part of the functions described above. In addition, the system and method embodiments provided in the above embodiments belong to the same concept, and their specific implementation process is detailed in the method embodiments, and will not be repeated here.

[0104] The above description is merely a preferred embodiment of the present invention and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of disclosure in this invention is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-disclosed concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions disclosed in this invention.

[0105] It should be noted that the terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and represent a limitation on a specific order or sequence. Where appropriate, the order of use for similar objects can be interchanged so that the embodiments of this application described herein can be implemented in an order other than that shown or described.

[0106] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A vision-guided multi-robotic arm collaborative coal cleaning system, characterized in that, include: The ground rail assembly includes four sets of ground rails arranged in parallel within the coal bunker below the tipper, with the installation direction of the four sets of ground rails perpendicular to the length direction of the open wagon. Four ground-rail mobile platforms are respectively installed on four sets of ground rails and move in a straight line along the ground rails; Four robotic arms are fixed to the four ground-rail moving platforms respectively, and each robotic arm is equipped with a high-pressure nozzle at its end. A high-pressure supply device, connected to the high-pressure nozzle, is used to supply high-pressure air to the high-pressure nozzle; An angle detection device is installed on the rotating shaft of the tippler to detect the tipping angle of the tippler. A lidar detection device is used to scan the inner wall of the open wagon to obtain three-dimensional point cloud data of the inner wall of the wagon. The control system is electrically connected to the ground-rail moving platform, the robotic arm, the high-voltage supply device, the angle detection device, and the lidar detection device; the control system is used for: Based on the three-dimensional point cloud data, the positions of the three sets of transverse struts inside the open wagon are identified, and the interior of the wagon is divided into four independent areas according to the positions of the three sets of struts. The four robotic arms are then bound to the four independent areas one by one. When the tipping angle reaches the preset trigger angle, the four ground rail moving platforms are controlled to drive the four robotic arms to extend synchronously from the initial position of the side wall of the coal bunker along the ground rail into the area below their respective independent areas. Based on the three-dimensional point cloud data, the residual coal distribution characteristics of each independent region are extracted, and a dedicated purging path matching the residual coal distribution characteristics of the corresponding independent region is generated for each robotic arm. The dedicated purging path is selected from at least one of the following: serpentine scanning path, fixed-point focusing path, and edge-attached path. The four robotic arms are controlled to perform high-pressure air purging along their respective dedicated purging paths, and the end-effector working spaces of adjacent robotic arms are constrained to not overlap at the boundaries defined by the three sets of support rods.

2. The vision-guided multi-robotic arm collaborative coal cleaning system according to claim 1, characterized in that, The serpentine scanning path involves the high-pressure nozzle at the end of the robotic arm reciprocating along the width of an independent region and gradually advancing along its length. The relationship between the scanning interval and the effective purging radius of the high-pressure nozzle satisfies... ,in For scanning spacing, To achieve an effective purging radius, This is the overlap coefficient.

3. The vision-guided multi-robotic arm collaborative coal cleaning system according to claim 1, characterized in that, The fixed-point focusing path involves the high-pressure nozzle at the end of the robotic arm moving to the coordinates of a point in a local area where the buildup thickness exceeds a preset threshold, then pausing for a preset time to perform enhanced purging. The relationship between the dwell time and the buildup thickness satisfies... ,in For the duration of stay, Minimum stay time, For time coefficient, The thickness of the deposit.

4. The vision-guided multi-robotic arm collaborative coal cleaning system according to claim 1, characterized in that, The edge-attaching path is defined as the high-pressure nozzle at the end of the robotic arm moving along the contour line of the side wall of the carriage or the edge line of the floor plate. A preset distance is maintained between the high-pressure nozzle and the wall surface, and a preset angle is maintained between the blowing direction and the normal to the wall surface. The relationship between the moving speed of the high-pressure nozzle and the residual coal adhesion strength level satisfies... ,in For movement speed, As the reference speed, The velocity attenuation coefficient, This refers to the adhesion strength rating.

5. The vision-guided multi-robotic arm collaborative coal cleaning system according to claim 1, characterized in that, The control system is also used for: When any of the robotic arms needs to extend across the strut boundary to an adjacent independent area for purging operations, a pause command is sent to the adjacent robotic arm to keep it in a retracted posture. After the robotic arm that crossed the boundary completes the operation and returns to its independent area, a resume command is sent to the adjacent robotic arm.

6. The vision-guided multi-robotic arm collaborative coal cleaning system according to claim 1, characterized in that, The ground rail is a triangular right-angled upward structure, which is suspended and installed on the bottom surface of the coal bunker by a bracket, and a preset distance is maintained between the ground rail and the bottom surface of the coal bunker.

7. A vision-guided multi-robotic arm collaborative coal cleaning system according to claim 6, characterized in that, The bottom of the ground rail moving platform is fixed with a wear-resistant scraper, which is in contact with the right-angle surface of the ground rail and is used to scrape off the coal material on the surface of the ground rail when the ground rail moving platform moves.

8. The vision-guided multi-robotic arm collaborative coal cleaning system according to claim 1, characterized in that, The control system is specifically used for: The start-up sequence of the four robotic arms is determined according to the purging duration of their respective independent areas. The robotic arm with the longer purging duration starts first, and the four robotic arms reset synchronously after purging is completed.

9. A vision-guided multi-robotic arm collaborative coal cleaning system according to claim 1, characterized in that, The lidar detection device is installed below the tipper or on the ground rail moving platform to scan the inner wall of the wagon before the robotic arm extends into the wagon.

10. A vision-guided multi-robotic arm collaborative method for cleaning residual coal, characterized in that, The method of using the vision-guided multi-robotic arm collaborative coal cleaning system as described in any one of claims 1 to 9 includes: Based on the three-dimensional point cloud data, the positions of the three sets of transverse struts inside the open wagon are identified, and the interior of the wagon is divided into four independent areas according to the positions of the three sets of struts. The four robotic arms are then bound to the four independent areas one by one. When the tipping angle reaches the preset trigger angle, the four ground rail moving platforms are controlled to drive the four robotic arms to extend synchronously from the initial position of the side wall of the coal bunker along the ground rail into the area below their respective independent areas. Based on the three-dimensional point cloud data, the residual coal distribution characteristics of each independent region are extracted, and a dedicated purging path matching the residual coal distribution characteristics of the corresponding independent region is generated for each robotic arm. The dedicated purging path is selected from at least one of the following: serpentine scanning path, fixed-point focusing path, and edge-attached path. The four robotic arms are controlled to perform high-pressure air purging along their respective dedicated purging paths, and the end-effector working spaces of adjacent robotic arms are constrained to not overlap at the boundaries defined by the three sets of support rods.

Images