Ore grabbing robot, control system and control method

Abstract

This invention relates to the technical field of robotic arms, proposing an ore-grabbing robotic arm, a control system, and a control method. The ore-grabbing robotic arm includes a moving device, a crossbeam base, a binocular camera, a sliding beam, a sliding beam drive component, a steering base, a steering drive component, a spacing slide, a spacing drive component, a lifting slide, a lifting drive component, a gripper mounting base, a gripper rotation drive component, a gripper camera, a mechanical gripper, a gripper drive component, and a control processor. By preprocessing the ore images captured by the binocular camera and establishing ore point clouds, followed by filtering and recognition, calculation and planning are performed to generate a gripping scheme. The optimal gripping scheme is obtained through scoring and optimization, and then the mechanical gripper is driven to operate. The feedback from the gripper camera is incorporated into the closed-loop control. The overall structure has higher ore-grabbing efficiency, and the optimal gripping scheme obtained through analysis provides more stable and reliable ore gripping.

Description

Technical Field

[0001] This invention relates to the technical field of robotic arms, and in particular to an ore-grabbing robotic arm, a control system, and a control method. Background Technology

[0002] In mining, ore transfer, and mineral processing operations, automated ore handling and loading are crucial for improving production efficiency and reducing labor intensity and safety risks. Currently, the industry commonly uses cranes equipped with grab buckets or robotic arms modified from construction machinery (such as excavators). However, when faced with ores that are irregularly shaped, vary in size, have complex surface characteristics, and exhibit random stacking patterns, existing technologies have significant shortcomings in terms of adaptability, precision, stability, and intelligence in handling operations.

[0003] Traditional grabs or fixed-track robotic arms have relatively fixed grasping action patterns and limited degrees of freedom. Most on-site operations still rely on the operator's visual judgment and manual control of the equipment to complete positioning and grasping. This method is inefficient, and the quality of work varies greatly depending on the operator's skill level and fatigue. Existing robotic arm systems mostly rely on single-view 2D cameras or coarse ranging sensors, which can only provide the planar position and approximate outline of the ore. They cannot obtain precise 3D dimensions, volume, center of gravity position, and other features. Without this key information, the control system cannot calculate the optimal gripping point, gripping force, and approach path. The grasping behavior is largely blind and trial-and-error, which can easily lead to missing the target area, gripping off-target, or collisions with non-target areas of the ore, causing damage to the equipment or the ore.

[0004] Most robotic arms currently employ single-jaw or double-jaw structures. Single-jaw grippers (such as grabs) primarily rely on gravity or insertion force to gather ore within the container, which is effective for loose materials. However, they cannot reliably grip large, solid, independent blocks of ore. While double-jaw grippers provide two points of force application, their gripping principle is essentially a force couple in a two-dimensional plane, providing constraint only in a single direction. For irregularly shaped ores in three-dimensional space, double-jaw grippers cannot restrict the rotational freedom of the ore around the line connecting the two grippers, easily leading to uncontrollable twisting and slippage during lifting, resulting in poor stability. Single-jaw grippers cannot form a force closure, relying entirely on friction and shape-based engagement, leading to low reliability. Double-jaw grippers only provide two contact points, and the conditions for forming a stable force closure in three-dimensional space are extremely demanding, typically requiring the contact points to be precisely located at specific positions on the friction cone, and the ore to have a highly regular shape. For most natural irregular ores, the double grippers have difficulty or even cannot find a pair of gripping points that can achieve stable force closure, forcing the system to excessively increase the gripping force to compensate, increasing the risk of crushing the ore or damaging the gripper teeth. Summary of the Invention

[0005] To address the aforementioned shortcomings, the present invention aims to provide an ore-grabbing robot, a control system, and a control method to solve the problems of low ore-grabbing efficiency and poor gripping stability in existing technologies.

[0006] To achieve this objective, the present invention adopts the following technical solution: A ore-grabbing robot, characterized in that it comprises a moving device, a crossbeam base, a binocular camera, a sliding beam, a sliding beam drive component, a steering base, a steering drive component, a spacing slide, a spacing drive component, a lifting slide, a lifting drive component, a gripper mounting base, a gripper rotation drive component, a gripper camera, a mechanical gripper, a gripper drive component, and a control processor; the crossbeam base is disposed at the output end of the moving device, and the moving device is used to move the crossbeam base to above the ore; the binocular camera is disposed at the center position of the crossbeam base; The sliding beams are provided in two sets, and the two sets of sliding beams are respectively slidably disposed at the left and right ends of the crossbeam seat. The two sets of sliding beam driving components respectively drive the two sets of sliding beams to slide along the horizontal X direction. The steering seat is disposed below the sliding beam, the steering seat is rotatably connected to the sliding beam, and the steering drive is used to drive the steering seat to rotate around the vertical axis; The bottom surface of the steering seat is slidably provided with two sets of the spacing slide blocks in the horizontal direction, and the spacing driving member is used to drive the two sets of the spacing slide blocks to slide linearly in the horizontal Y direction. The inner side of the spacing slide is slidably provided with the lifting slide in the vertical direction; the lifting drive is used to drive the lifting slide to slide linearly in the vertical Z direction. The lifting slide is rotatably equipped with the gripper mounting seat, and the gripper rotation drive is used to drive the gripper mounting seat to rotate along the vertical axis; The four sets of gripper mounting bases are respectively equipped with mechanical grippers and gripper cameras; the gripper drive is used to drive the mechanical grippers to open or close. The binocular camera and the four sets of gripper cameras are respectively connected to the control processor via signals.

[0007] Preferably, the bottom of the sliding beam is provided with several sets of 360° circular slide rails, the diameter of each set of circular slide rails gradually increases, the steering seat is cylindrical, and the top of the steering seat is provided with several sets of annular slider groups corresponding to the diameter of the circular slide rails. Each set of annular slider groups includes multiple sliders arranged in a circle, and the sliders are slidably connected to the corresponding circular slide rails.

[0008] Preferably, the sliding beam is provided with a through sliding hole, the sliding hole is sleeved on the outer periphery of the crossbeam seat, the upper end surface of the crossbeam seat is provided with a support guide rail, and the top of the sliding hole is provided with a support slider, which is slidably disposed on the support guide rail.

[0009] This invention proposes a control system for an ore-grabbing robot, which is applied to the ore-grabbing robot and includes an image preprocessing module, a 3D modeling module, a point cloud recognition module, a gripping point planning module, a scoring optimization module, a control drive module, and a feedback adjustment module. The image preprocessing module is used for preprocessing images to perform noise reduction and image enhancement. The 3D modeling module is used to recognize and generate the original point cloud model from the preprocessed image, and to perform noise reduction and filtering on the original point cloud model to obtain the 3D point cloud of the ore. The point cloud recognition module is used to analyze the shape and size of the ore point cloud based on the ore's 3D point cloud. The clamping point planning module is used to calculate the clamping points of the four mechanical grippers to obtain a clamping planning set; The scoring and optimization module is used to score and optimize each set in the clamping plan set to obtain the ore coordinate system of the four clamping points of the optimal clamping plan; The control drive module is used to calculate and convert control commands based on the ore coordinate system of the four clamping points and output control commands to the drive components. The feedback adjustment module is used to receive images from the gripper camera and make real-time feedback adjustments based on the images.

[0010] This invention proposes a control method for an ore-grabbing robot, applied to the ore-grabbing robot, comprising: Step S1: Preprocessing the image by denoising and enhancing it; Step S2: Recognize the two-dimensional contour of the ore from the preprocessed image, generate the original point cloud model by combining the parallax principle of the depth camera, perform noise reduction and filtering on the original point cloud model, remove background interference points, and obtain the three-dimensional point cloud of the ore. Step S3: Analyze the shape and size of the ore point cloud based on the 3D point cloud analysis; Step S4: Based on the shape and size of the ore point cloud, calculate the clamping points of the four mechanical grippers 10 to obtain the clamping plan set; Step S5: Score and optimize each set in the clamping plan set to obtain the ore coordinate system of the four clamping points of the optimal clamping plan. ; Step S6: Calculate and control the ore coordinate system of the four clamping points according to the optimal clamping plan, and output control commands to drive the four mechanical jaws to move to the four optimal clamping points obtained by the optimal clamping plan through the drive components. Step S7: Receive the image from the gripper camera and provide real-time feedback to adjust the position of the mechanical gripper and clamp the ore.

[0011] Preferably, step 2 includes the following method: applying a Canny edge detector to the enhanced image to identify the two-dimensional contour of the ore; Using two views from a binocular camera, a semi-global block matching algorithm is used to calculate the disparity value for each pixel. Combined with the known camera focal length and baseline distance, the disparity is converted into the depth of each pixel based on the principle of triangulation. Based on the camera's intrinsic and extrinsic parameters, geometric formulas are used to convert the coordinates of each pixel into a 3D point in the world coordinate system, forming the original point cloud model. The original point cloud model is then statistically removed to eliminate floating and unreasonable stray points.

[0012] Preferably, step S5 includes: A scoring model is established, which includes stability score, balance score, safety score, and exercise efficiency score; the comprehensive scoring function is: ,in: The overall score ranges from 0 to 1. Let be the weighting coefficient, satisfying ; Assess stability score; For balance scoring; For safety rating, Scoring of exercise efficiency; A comprehensive score is calculated for each set in the clamping plan set, and the scheme with the highest score is selected as the optimal clamping plan.

[0013] Preferred stability score , To achieve the minimum required clamping force, This represents the maximum clamping force of the grippers, where W is the weight of the ore, K is the safety factor, and μ is the coefficient of friction; Balance score ,in The maximum clamping force difference of the four grippers ; The average clamping force of the four grippers. , Let be the distance from the i-th clamping point to the center of gravity of the ore; Security Score , This is the minimum distance between the gripper and the non-gripping area of ​​the ore. The set safety threshold; Sports efficiency score , The total time to complete the clamping action is t, which is the reference time base.

[0014] Preferably, step S6 further includes the following steps: Step S6-1: Based on the optimal clamping plan, determine the clamping points. The ore coordinate system is transformed into the world coordinate system using a transformation matrix; Step S6-2: Perform inverse kinematics solution based on the world coordinate system to calculate the target positions of the sliding beam drive, spacing drive, lifting drive and gripper drive, as well as the target angles of the steering drive and gripper rotation drive. Step S6-3: Based on the target positions of the slide beam drive, spacing drive, lifting drive and gripper drive, and the target angles of the steering drive and gripper rotation drive, plan the path of each drive component; Step S6-4: Based on the planned path, convert the physical quantities into servo control parameters and send them to the slide beam drive, spacing drive, lifting drive, gripper drive, steering drive and gripper rotation drive for drive control.

[0015] Preferably, step S4 includes generating clamping candidate points based on the shape and size of the ore point cloud, generating a clamping point candidate set from the clamping candidate points through a constraint model, and performing a rapid geometric screening from the clamping point candidate set to eliminate obviously unreasonable combinations.

[0016] One of the above technical solutions has the following advantages or beneficial effects: 1. The ore gripping robot achieves overall movement, lateral extension and retraction, spacing adjustment, independent lifting and rotation through a moving device, sliding beam, spacing slide, lifting slide, steering seat and gripper mounting seat. It is decomposed into a series of orderly simple movements, which reduces the load and motion complexity of a single drive component, improves the reliability and life of the overall system, and makes its working space and posture adjustment capabilities far exceed those of traditional fixed or single-arm robots. It can flexibly surround and conform to the irregular surface of the ore like a human hand and find the optimal gripping point. 2. The four mechanical grippers can be independently controlled in position and attitude. Through intelligent planning, they can form a spatially stable closed grip, applying clamping force from multiple directions simultaneously to firmly lock the ore in the center. This is more reliable than common single-point hoisting or two-point clamping and reduces the risk of ore swinging and collision.

[0017] 3. Each mechanical gripper is independently equipped with a gripper camera. When the gripper approaches the ore, it provides a first-person perspective ultra-close-up image to accurately identify the micron-level positional deviation between the gripper and the predetermined gripping point. This information is fed back to the control system in real time for fine-tuning of the end-effector's posture, ensuring that the gripper accurately lands on the optimal gripping point.

[0018] 4. The control method uses image preprocessing and point cloud filtering to resist changes in illumination and minor noise interference. It plans and generates and optimizes geometric models with uncertainties, plans the optimal clamping scheme, and then drives the mechanical gripper to operate. By incorporating the feedback from the gripper camera into the closed loop, final fine-tuning and adaptive adjustments can be made during the contact stage, forming an enhanced closed loop of "perception-planning-action-re-perception". The overall structure has higher ore clamping efficiency, and the optimal clamping scheme obtained through analysis provides more stable and reliable ore clamping. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the overall structure of the ore-grabbing robot proposed in this invention. Figure 2 This is a partially enlarged schematic diagram of the ore-grabbing robot proposed in this invention; Figure 3 This is a flowchart of the ore-grabbing manipulator control method proposed in this invention; Figure 4 This is a flowchart of step S6 of the ore-grabbing manipulator control method proposed in this invention; The components include: 1. mobile device; 2. crossbeam seat; 3. binocular camera; 4. sliding beam; 5. steering seat; 6. spacing slide; 7. lifting slide; 8. gripper mounting seat; 9. gripper camera; 10. mechanical gripper; 11. control processor; 41. sliding hole; and 21. support guide rail. Detailed Implementation

[0020] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0021] In the description of this invention, it should be understood that the terms "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "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 used only for the convenience of describing the invention and for 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 invention. Furthermore, features defined with "first" and "second" may explicitly or implicitly include one or more of these features, used to distinguish and describe features, without any order or emphasis.

[0022] In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0023] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0024] The following is combined with Figures 1 to 4 This invention describes an ore-grabbing robot, its control system, and its control method according to embodiments of the present invention. One type of ore-grabbing robot includes a moving device 1, a crossbeam seat 2, a binocular camera 3, a sliding beam 4, a sliding beam drive component, a steering seat 5, a steering drive component, a spacing slide 6, a spacing drive component, a lifting slide 7, a lifting drive component, a gripper mounting seat 8, a gripper rotation drive component, a gripper camera 9, a mechanical gripper 10, a gripper drive component, and a control processor 11; the crossbeam seat 2 is disposed at the output end of the moving device 1, and the moving device 1 is used to drive the crossbeam seat 2 to move above the ore; the binocular camera 3 is disposed at the center position of the crossbeam seat 2; Two sets of sliding beams 4 are provided, and the two sets of sliding beams 4 are respectively slidably disposed at the left and right ends of the crossbeam seat 2. The two sets of sliding beam driving components respectively drive the two sets of sliding beams 4 to slide along the horizontal X direction. The steering seat 5 is disposed below the sliding beam 4, and the steering seat 5 is rotatably connected to the sliding beam 4. The steering drive component is used to drive the steering seat 5 to rotate around the vertical axis. The bottom surface of the steering seat 5 is slidably provided with two sets of the spacing slide blocks 6 in the horizontal direction, and the spacing driving member is used to drive the two sets of the spacing slide blocks 6 to slide linearly in the horizontal Y direction. The inner side of the spacing slide 6 is slidably provided with the lifting slide 7 in the vertical direction; the lifting drive is used to drive the lifting slide 7 to slide linearly in the vertical Z direction. The lifting slide 7 is rotatably provided with the gripper mounting seat 8, and the gripper rotation drive is used to drive the gripper mounting seat 8 to rotate along the vertical axis. The four sets of gripper mounting bases 8 are respectively provided with mechanical grippers 10 and gripper cameras 9; the gripper driving component is used to drive the mechanical grippers 10 to open or close. The binocular camera 3 and the four sets of gripper cameras 9 are respectively connected to the control processor 11 via signal.

[0025] The specific steps for gripping the ore are as follows: Based on the ore position captured by the binocular camera 3, the optimal gripping scheme is calculated, and then the gripper is driven to move forward to grip the ore. Specifically: The moving device 1 moves the crossbeam seat 2 to directly above the ore and lowers it to a safe distance. Then, the steering drive component drives the steering seat 5 to rotate to the controlled angle. The sliding beam drive component drives the crossbeam seat 2 to move to the left and right sides of the ore, maintaining a safe distance to ensure that it will not collide with the ore during descent. After that, the gripper drive component drives the mechanical gripper to open, and the moving device 1 moves the crossbeam seat 2 to a precise positioning height, generally the ore. Near the center, the lifting drive unit drives the four lifting seats to move to the corresponding gripper height, ensuring that all four mechanical grippers 10 can grip the ore. Then, the gripper rotation drive unit drives the gripper mounting base 8 to rotate, ensuring that the mechanical grippers 10 face the ore for gripping. Next, the spacing drive unit drives the two spacing slides 6 to move to the accurate gripping width. The slide beam drive unit drives the crossbeam base 2 to move closer to the gripping length. This position of the robot arm is the optimal gripping position. Then, the gripper drive unit drives the mechanical gripper to clamp and output the corresponding gripping force to tighten the ore.

[0026] The mobile device 1 can be a robotic arm or a mobile gantry machine tool, etc. Different mobile devices 1 can be used in different application scenarios. The crossbeam seat 2 can be installed at the output end of the mobile device 1. This invention realizes overall movement, lateral extension and retraction, spacing adjustment, independent lifting and independent rotation through the mobile device 1, sliding beam 4, spacing slide 6, lifting slide 7, steering seat 5 and gripper mounting seat 8. The mobile device 1 is responsible for large-range positioning, the sliding beam 4 and spacing slide 6 are responsible for lateral envelopment, the lifting slide 7 is responsible for height matching, and the steering seat 5 and gripper mounting seat 8 are responsible for angle alignment. This division of labor decomposes the complex grasping task into a series of orderly simple movements, reduces the load and motion complexity of a single drive component, improves the reliability and lifespan of the overall system, and makes its workspace and posture adjustment capabilities far exceed those of traditional fixed or single-arm robotic arms. It can flexibly surround and conform to the irregular surface of the ore like a human hand to find the optimal gripping point, rather than forcing the ore to adapt to a fixed gripping path; the four mechanical grippers 10 can independently control their position and posture. Through intelligent planning, they can form a spatially stable closed gripping mechanism, applying clamping forces from multiple directions simultaneously to firmly lock the ore in the center. This is more reliable than common single-point hoisting or two-point clamping, reducing the risk of ore swinging and colliding.

[0027] Each mechanical gripper 10 is independently equipped with a gripper camera 9, providing a first-person, ultra-close-up image as the gripper approaches the ore. This camera accurately identifies micron-level positional deviations between the gripper and the predetermined gripping point, providing real-time feedback to the control system for fine-tuning the end effector's posture. This ensures the gripper accurately lands at the optimal gripping point. The gripper camera 9 continues to operate during and after gripping, providing real-time visual feedback on the gripping status. For example, it can monitor whether the contact surface between the gripper and the ore is properly aligned and whether there are any signs of slippage. Based on this feedback, the control processor 11 can fine-tune the gripping force or angle of the gripper at the moment of gripping, achieving adaptive gripping.

[0028] Furthermore, the bottom of the sliding beam 4 is provided with several sets of 360° circular slide rails, the diameter of each set of circular slide rails gradually increases, the steering seat 5 is cylindrical, and the top of the steering seat 5 is provided with several sets of annular slider groups corresponding to the diameter of the circular slide rails. Each set of annular slider groups includes multiple sliders arranged in a circle, and the sliders are slidably connected to the corresponding circular slide rails.

[0029] Specifically, several sets of circular slide rails with gradually increasing diameters work together with corresponding multi-slider annular slider groups to form a composite support surface radiating outwards from the center. This effectively distributes the axial force, radial force, and overturning moment acting on the steering seat 5 to a larger area of ​​the sliding beam 4 through slide rail layers of different diameters, significantly improving bending and torsional stiffness. Especially when gripping ore whose center of gravity deviates from the axis of the robotic arm, it effectively suppresses the slight tilting of the steering seat 5, ensuring precise posture. Multiple sets of evenly arranged circumferential sliders and circular slide rails form a multi-point uniform constraint. Regardless of the angle of the steering seat 5, the load can be evenly transmitted through multiple sliders, avoiding jamming caused by uneven force distribution. This ensures the smoothness and high precision of the rotational motion, providing a foundation for the precise alignment of the gripper. The 360° circumferential arrangement of slide rails and slider assemblies forms a closed guide and load-bearing ring. No matter which direction the lateral force comes from, there is a corresponding contact surface between the slider assembly and the slide rail to bear it, providing all-round, no-dead-angle rigid support. This greatly enhances the structure's adaptability to dynamic loads, reduces the direct impact of impact on the steering drive motor and reducer, and thus extends the service life of the entire rotating unit.

[0030] Furthermore, the sliding beam 4 is provided with a through sliding hole 41, which is sleeved on the outer periphery of the crossbeam seat 2. The upper end face of the crossbeam seat 2 is provided with a support guide rail 21, and a support slider is provided inside the top of the sliding hole 41. The support slider is slidably mounted on the support guide rail 21.

[0031] Specifically, in traditional suspended or single-sided guide rail designs, the sliding beam 4 is essentially suspended on one side of the crossbeam seat 2, with its center of gravity outside the guide rail, creating an unstable torque. This embodiment features a sliding hole 41, allowing the crossbeam seat 2 to completely pass through the sliding beam 4, forming a full-circumferential enclosure constraint. This effectively prevents the sliding beam 4 from twisting around the crossbeam seat 2. A support guide rail 21 is provided on the upper surface of the crossbeam seat 2, cooperating with the support slider at the top of the sliding hole 41. The support guide rail 21 is located on the upper surface of the crossbeam seat 2, at the position with the best bending resistance in the cross-section of the crossbeam seat 2. The load is transmitted directly to the upper surface of the crossbeam seat 2 via the sliding beam 4 → top support slider → support guide rail 21, resulting in extremely strong resistance to overturning moments. The support slider and guide rail bear almost all the vertical downward load, while the gap between the inner wall of the sliding hole 41 and the outer periphery of the crossbeam seat 2 mainly serves as radial positioning and anti-torsion, bearing a small radial force. This avoids the problem of high friction and rapid wear caused by a single contact surface having to bear the load and guide. The top support can use a heavy-duty linear guide rail or a hydrostatic slider, which is specially optimized for high load-bearing and low-friction linear motion. The sliding hole 41 can use a low-friction coefficient material bushing or maintain a large clearance fit, requiring only a small force to overcome radial friction. The overall driving resistance is significantly reduced, and the motion is smoother.

[0032] The present invention proposes a control system for an ore gripping robot, which is applied to the ore gripping robot and includes an image preprocessing module, a 3D modeling module, a point cloud recognition module, a gripping point planning module, a scoring optimization module, a control drive module, and a feedback adjustment module. The image preprocessing module is used for preprocessing images to perform noise reduction and image enhancement. The 3D modeling module is used to recognize and generate the original point cloud model from the preprocessed image, and to perform noise reduction and filtering on the original point cloud model to obtain the 3D point cloud of the ore. The point cloud recognition module is used to analyze the shape and size of the ore point cloud based on the ore's 3D point cloud. The clamping point planning module is used to calculate the clamping points of the four mechanical grippers 10 to obtain a clamping planning set; The scoring and optimization module is used to score and optimize each set in the clamping plan set to obtain the ore coordinate system of the four clamping points of the optimal clamping plan; The control drive module is used to calculate and convert control commands based on the ore coordinate system of the four clamping points and output control commands to the drive components. The feedback adjustment module is used to receive the image from the gripper camera 9 and make real-time feedback adjustments based on the image.

[0033] The present invention proposes a control method for an ore-grabbing manipulator, applied to the ore-grabbing manipulator, comprising: Step S1: Preprocessing the image by denoising and enhancing it; Step S2: Recognize the two-dimensional contour of the ore from the preprocessed image, generate the original point cloud model by combining the parallax principle of the depth camera, perform noise reduction and filtering on the original point cloud model, remove background interference points, and obtain the three-dimensional point cloud of the ore. Step S3: Analyze the shape and size of the ore point cloud based on the 3D point cloud analysis; Step S4: Based on the shape and size of the ore point cloud, calculate the clamping points of the four mechanical grippers 10 to obtain the clamping plan set; Step S5: Score and optimize each set in the clamping plan set to obtain the ore coordinate system of the four clamping points of the optimal clamping plan. ; Step S6: Calculate and control the ore coordinate system of the four clamping points according to the optimal clamping plan, and output control commands to drive the four mechanical grippers 10 to move to the four optimal clamping points obtained by the optimal clamping plan through the drive components. Step S7: Receive the image from the gripper camera 9 and provide real-time feedback to adjust the position of the mechanical gripper 10 and clamp the ore.

[0034] Specifically, step S2 provides the control system with a high-precision 3D point cloud of the ore in the current field of view through stereo vision and depth calculation. The model contains all key geometric information such as the ore's true shape, size, surface undulations, and spatial orientation, providing a unique and accurate input for subsequent intelligent analysis. Steps S3 and S4 constitute the core cognitive layer. The algorithm automatically extracts features such as bounding box size, volume, center of gravity, principal curvature of the surface, and concave and convex regions from the point cloud. Based on these features, it automatically generates a series of feasible four-point clamping schemes using grasping mechanics and geometric constraints. Step S5 establishes a quantitative scoring model to comprehensively evaluate and rank each candidate plan in the set, ensuring that the final executed scheme is the globally optimal solution after comprehensive consideration, thereby fundamentally improving the grasping performance. Reliability, safety, and economy; Step S6 involves complex coordinate transformations and inverse kinematics solutions. The system needs to transform the optimal gripping point in the ore coordinate system through a series of coordinate transformations to finally calculate the distance or angle of movement required for each sliding beam drive component, steering drive component, spacing drive component, lifting drive component, and gripper rotation drive component. This is the foundation for the complex robot to accurately execute the plan. The entire process has inherent robustness. The image preprocessing and point cloud filtering in S1 and S2 can resist changes in illumination and a small amount of noise interference. The planning generation process S4 and optimization process S5 handle the geometric model containing uncertainties. Finally, by incorporating the feedback from the gripper camera 9 into the closed loop, the final fine-tuning and adaptation can be performed in the contact stage, forming an enhanced closed loop of "perception-planning-action-re-perception".

[0035] Furthermore, step 2 includes the following method: applying a Canny edge detector to the enhanced image to identify the two-dimensional contours of the ore; Using the two views of the binocular camera 3, a semi-global block matching algorithm is used to calculate the disparity value of each pixel. Combined with the known camera focal length and baseline distance, the disparity is converted into the depth of each pixel according to the principle of triangulation. Based on the camera intrinsic and extrinsic parameters, the coordinates of each pixel are converted into a three-dimensional point in the world coordinate system using geometric formulas to form the original point cloud model. The original point cloud model is then statistically removed to eliminate floating and unreasonable stray points.

[0036] Furthermore, step S5 includes: A scoring model is established, which includes stability score, balance score, safety score, and exercise efficiency score; the comprehensive scoring function is: ,in: The overall score ranges from 0 to 1. Let be the weighting coefficient, satisfying ; Assess stability score; For balance scoring; For safety rating, Scoring of exercise efficiency; A comprehensive score is calculated for each set in the clamping plan set, and the scheme with the highest score is selected as the optimal clamping plan.

[0037] Furthermore, stability score , To achieve the minimum required clamping force, This represents the maximum clamping force of the grippers, where W is the weight of the ore, K is the safety factor, and μ is the coefficient of friction; Balance score ,in The maximum clamping force difference of the four grippers ; The average clamping force of the four grippers. , Let be the distance from the i-th clamping point to the center of gravity of the ore; Security Score , This is the minimum distance between the gripper and the non-gripping area of ​​the ore. The set safety threshold; Sports efficiency score , The total time to complete the clamping action is t, which is the reference time base.

[0038] Specifically, by transforming the scoring model from a "feasible solution" to an "optimal solution," not only is the blindness of random or empirical selection avoided, but the quantitative evaluation system also enables the grasping strategy to be more scientific, precise, and predictable. The force closure test theoretically guarantees the static stability of the gripping mechanism, preventing slippage or rotation during grasping. The stability margin quantitatively assesses the ability to resist disturbances, ensuring stability even under actual vibration, wind load, and other disturbances. Physical feasibility checks eliminate solutions that appear feasible but may actually lead to collisions or exceed limits. This significantly improves the gripping success rate and stability. The balance scoring mandates uniform force distribution across the four grippers, avoiding single-point overload. Traditional methods may result in one gripper bearing more than 50% of the load while others bear only a small portion. After optimization, the scoring system ensures that each gripper... The gripper load variation is controlled within a manageable range. The motion efficiency score selects the scheme with the shortest total motion distance and smoothest acceleration, reducing motor start-stop impact, lowering peak current requirements, and optimizing mechanical system lifespan and energy consumption. The safety score directly quantifies the distance between the gripper and the ore surface, mandating a minimum safe distance (e.g., 30mm). Schemes with insufficient distance are penalized or eliminated. The stability margin provides a safety buffer; even if the actual friction coefficient is lower than estimated, or even if there are minor errors in the ore's center of gravity calculation, the system can still maintain stable gripping. The system can automatically optimize scoring weights based on historical gripping data, identify and avoid repetitive failure patterns, and establish a "ore characteristics - optimal gripping strategy" knowledge base. Each gripping operation records complete scoring data and actual results. Through machine learning methods, weight parameters are optimized and adjusted... The proportion of scores is used to identify the combination of scoring features that leads to failure, record learning failures, build a predictive model for new ores, and quickly recommend possible optimal parameters.

[0039] Furthermore, step S6 also includes the following steps: Step S6-1: Based on the optimal clamping plan, determine the clamping points. The ore coordinate system is transformed into the world coordinate system using a transformation matrix; Step S6-2: Perform inverse kinematics solution based on the world coordinate system to calculate the target positions of the sliding beam drive, spacing drive, lifting drive and gripper drive, as well as the target angles of the steering drive and gripper rotation drive. Step S6-3: Based on the target positions of the slide beam drive, spacing drive, lifting drive and gripper drive, and the target angles of the steering drive and gripper rotation drive, plan the path of each drive component; Step S6-4: Based on the planned path, convert the physical quantities into servo control parameters and send them to the slide beam drive, spacing drive, lifting drive, gripper drive, steering drive and gripper rotation drive for drive control.

[0040] Specifically, the optimal gripping plan is generated in the ore coordinate system, while the motion control of each joint of the manipulator must be performed in a unified world coordinate system. Through a precisely calibrated transformation matrix containing the position and orientation information of the ore, coordinate transformation is performed to map the planned points to an absolute spatial position that the manipulator can understand and execute. This ensures that the final spatial position reached by the gripper precisely matches the planned position of the ore surface to be gripped, eliminating gripping deviations caused by inconsistencies in coordinate systems. For the multi-degree-of-freedom manipulator, its inverse kinematics may not have an analytical solution or may have multiple solutions. Step S6-2 uses a numerical algorithm to calculate the precise target positions of each sliding beam, the spacing slide 6, and the lifting slide 7, as well as the steering seat 5 and the gripper mounting position, based on the target position and orientation of the gripper in the world coordinate system. The target angle that the mounting 8 needs to rotate is automatically and accurately decomposed into simple linear or rotational motions of multiple drive motors based on the three-dimensional target pose of the four grippers. After path planning in step 6-3, the drive components can move synchronously and in a coordinated manner, so that the four grippers reach the target point simultaneously along a smooth trajectory, avoiding disjointed movements, mutual interference, or unnecessary stops. The robot arm moves smoothly and without sudden starts and stops, reducing impact loads, mechanical wear, and vibration. In step S6-4, based on the mechanical transmission parameters of each drive component, such as lead screw, reduction ratio, and encoder resolution, the target position or angle is converted into the target pulse count or analog set value of the servo motor. Subsequently, these instructions are sent synchronously and promptly to each driver for simultaneous driving via a real-time control bus (or pulse train).

[0041] Furthermore, step S4 includes generating clamping candidate points based on the shape and size of the ore point cloud, generating a clamping candidate set from the clamping candidate points through a constraint model, and performing a rapid geometric screening from the clamping candidate set to eliminate obviously unreasonable combinations.

[0042] Specifically, before entering the computationally intensive scoring and optimization stage, points that the gripper cannot reach, have unreasonable postures, or poor contact are quickly eliminated. This prevents the optimization algorithm from getting lost or trapped in local optima among a large number of invalid solutions, making the system output more stable and predictable.

[0043] Other configurations and operations of the ore-grabbing robot, control system, and control method according to embodiments of the present invention are known to those skilled in the art and will not be described in detail here.

[0044] In the description of this specification, references to terms such as "embodiment," "example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0045] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A ore-grabbing robotic arm, characterized in that: The device includes a moving device, a crossbeam base, a binocular camera, a sliding beam, a sliding beam drive component, a steering seat, a steering drive component, a spacing slide, a spacing drive component, a lifting slide, a lifting drive component, a gripper mounting base, a gripper rotation drive component, a gripper camera, a mechanical gripper, a gripper drive component, and a control processor. The crossbeam base is located at the output end of the moving device, which is used to move the crossbeam base above the ore. The binocular camera is located at the center of the crossbeam base. The sliding beams are provided in two sets, and the two sets of sliding beams are respectively slidably disposed at the left and right ends of the crossbeam seat. The two sets of sliding beam driving components respectively drive the two sets of sliding beams to slide along the horizontal X direction. The steering seat is disposed below the sliding beam, the steering seat is rotatably connected to the sliding beam, and the steering drive is used to drive the steering seat to rotate around the vertical axis; The bottom surface of the steering seat is slidably provided with two sets of the spacing slide blocks in the horizontal direction, and the spacing driving member is used to drive the two sets of the spacing slide blocks to slide linearly in the horizontal Y direction. The inner side of the spacing slide is slidably provided with the lifting slide in the vertical direction; the lifting drive is used to drive the lifting slide to slide linearly in the vertical Z direction. The lifting slide is rotatably equipped with the gripper mounting seat, and the gripper rotation drive is used to drive the gripper mounting seat to rotate along the vertical axis; The four sets of gripper mounting bases are respectively equipped with mechanical grippers and gripper cameras; the gripper drive is used to drive the mechanical grippers to open or close. The binocular camera and the four sets of gripper cameras are respectively connected to the control processor via signals.

2. The ore-grabbing robotic arm according to claim 1, characterized in that: The bottom of the sliding beam is provided with several sets of 360° circular slide rails, the diameter of each set of circular slide rails gradually increases, the steering seat is cylindrical, and the top of the steering seat is provided with several sets of annular sliders corresponding to the diameter of the circular slide rails. Each set of annular sliders includes multiple sliders arranged in a circle, and the sliders are slidably connected to the corresponding circular slide rails.

3. The ore-grabbing robotic arm according to claim 1, characterized in that: The sliding beam has a through sliding hole, which is fitted around the outer periphery of the crossbeam seat. A support guide rail is provided on the upper end face of the crossbeam seat, and a support slider is provided inside the top of the sliding hole. The support slider is slidably mounted on the support guide rail.

4. A control system for an ore-grabbing robotic arm, characterized in that: The ore-grabbing robot as described in any one of claims 1-3 includes an image preprocessing module, a 3D modeling module, a point cloud recognition module, a gripping point planning module, a scoring optimization module, a control drive module, and a feedback adjustment module. The image preprocessing module is used for preprocessing images to perform noise reduction and image enhancement. The 3D modeling module is used to recognize and generate the original point cloud model from the preprocessed image, and to perform noise reduction and filtering on the original point cloud model to obtain the 3D point cloud of the ore. The point cloud recognition module is used to analyze the shape and size of the ore point cloud based on the ore's 3D point cloud. The clamping point planning module is used to calculate the clamping points of the four mechanical grippers to obtain a clamping planning set; The scoring and optimization module is used to score and optimize each set in the clamping plan set to obtain the ore coordinate system of the four clamping points of the optimal clamping plan; The control drive module is used to calculate and convert control commands based on the ore coordinate system of the four clamping points and output control commands to the drive components. The feedback adjustment module is used to receive images from the gripper camera and make real-time feedback adjustments based on the images.

5. A control method for an ore-grabbing robotic arm, characterized in that: The ore-grabbing robot as described in any one of claims 1-3 includes: Step S1: Preprocessing the image by denoising and enhancing it; Step S2: Recognize the two-dimensional contour of the ore from the preprocessed image, generate the original point cloud model by combining the parallax principle of the depth camera, perform noise reduction and filtering on the original point cloud model, remove background interference points, and obtain the three-dimensional point cloud of the ore. Step S3: Analyze the shape and size of the ore point cloud based on the 3D point cloud analysis; Step S4: Based on the shape and size of the ore point cloud, calculate the clamping points of the four mechanical grippers 10 to obtain the clamping plan set; Step S5: Score and optimize each set in the clamping plan set to obtain the ore coordinate system of the four clamping points of the optimal clamping plan. ; Step S6: Calculate and control the ore coordinate system of the four clamping points according to the optimal clamping plan, and output control commands to drive the four mechanical jaws to move to the four optimal clamping points obtained by the optimal clamping plan through the drive components. Step S7: Receive the image from the gripper camera and provide real-time feedback to adjust the position of the mechanical gripper and clamp the ore.

6. The control method for an ore-grabbing robotic arm according to claim 5, characterized in that: Step 2 includes the following method: applying a Canny edge detector to the enhanced image to identify the two-dimensional contours of the ore; Using two views from a binocular camera, a semi-global block matching algorithm is used to calculate the disparity value for each pixel. Combined with the known camera focal length and baseline distance, the disparity is converted into the depth of each pixel based on the principle of triangulation. Based on the camera's intrinsic and extrinsic parameters, geometric formulas are used to convert the coordinates of each pixel into a 3D point in the world coordinate system, forming the original point cloud model. The original point cloud model is then statistically removed to eliminate floating and unreasonable stray points.

7. The control method for an ore-grabbing robotic arm according to claim 6, characterized in that: Step S5 includes: A scoring model is established, which includes stability score, balance score, safety score, and exercise efficiency score; the comprehensive scoring function is: ,in: The overall score ranges from 0 to 1. Let be the weighting coefficient, satisfying ; Assess stability score; For balance scoring; For safety rating, Scoring of exercise efficiency; A comprehensive score is calculated for each set in the clamping plan set, and the scheme with the highest score is selected as the optimal clamping plan.

8. The control method for an ore-grabbing robotic arm according to claim 7, characterized in that: Stability rating , To achieve the minimum required clamping force, This represents the maximum clamping force of the grippers, where W is the weight of the ore, K is the safety factor, and μ is the coefficient of friction; Balance score ,in The maximum clamping force difference of the four grippers ; The average clamping force of the four grippers. , Let be the distance from the i-th clamping point to the center of gravity of the ore; Security Score , This is the minimum distance between the gripper and the non-gripping area of ​​the ore. The set safety threshold; Sports efficiency score , The total time to complete the clamping action is t, which is the reference time base.

9. The control method for an ore-grabbing robot according to claim 5, characterized in that: Step S6 further includes the following steps: Step S6-1: Based on the optimal clamping plan, determine the clamping points. The ore coordinate system is transformed into the world coordinate system using a transformation matrix; Step S6-2: Perform inverse kinematics solution based on the world coordinate system to calculate the target positions of the sliding beam drive, spacing drive, lifting drive and gripper drive, as well as the target angles of the steering drive and gripper rotation drive. Step S6-3: Based on the target positions of the slide beam drive, spacing drive, lifting drive and gripper drive, and the target angles of the steering drive and gripper rotation drive, plan the path of each drive component; Step S6-4: Based on the planned path, convert the physical quantities into servo control parameters and send them to the slide beam drive, spacing drive, lifting drive, gripper drive, steering drive and gripper rotation drive for drive control.

10. The control method for an ore-grabbing robot according to claim 5, characterized in that: Step S4 includes generating clamping candidate points based on the shape and size of the ore point cloud, generating a clamping candidate set from the clamping candidate points through a constraint model, and performing a rapid geometric screening from the clamping candidate set to eliminate obviously unreasonable combinations.

Images