Automatic large ball crushing device for disc pelletizer

The automatic large ball crushing device for disc pelletizing machines, which works in collaboration with an image acquisition module and a robotic arm, solves the problem of large balls affecting pelletizing quality and equipment operation, achieving efficient and accurate automated crushing, and improving production efficiency and equipment lifespan.

CN120155288BActive Publication Date: 2026-11-10BEIJING ZHONGHONGLIAN ENG TECH CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202510296613.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-03-13
Publication Date
2026-11-10
Estimated Expiration
2045-03-13

AI Technical Summary

Technical Problem

Existing disc pelletizers produce large pellets during the pelletizing process, which affects pelletizing quality and equipment operation. Furthermore, manual handling is inefficient and inaccurate, and market equipment is insufficient in terms of detection accuracy, crushing effect, and adaptability.

Method used

The system employs the collaborative operation of an image acquisition module, a control unit, a six-degree-of-freedom robotic arm, and a crusher. It identifies large balls through image analysis algorithms, plans obstacle avoidance trajectories through three-dimensional path planning algorithms, and uses the six-degree-of-freedom robotic arm to drive the crusher to precisely crush the large balls.

Benefits of technology

It enables automatic detection and crushing of large pellets, improves the automation and efficiency of pellet production, reduces manual intervention, lowers labor intensity, and improves pellet quality and equipment lifespan.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120155288B_ABST
    Figure CN120155288B_ABST
Patent Text Reader

Abstract

This invention discloses an automatic large ball crushing device for a disc pelletizing machine, belonging to the technical field of pelletizing equipment. In the disc pelletizing process, large balls often affect pelletizing quality and subsequent production processes. This device includes an image acquisition module for collecting image data of the pelletizing disc area, a control unit with built-in image analysis and 3D path planning algorithms, a six-degree-of-freedom robotic arm located near the pelletizing disc area, and a crushing hand mounted at the end of the robotic arm. The image analysis algorithm detects the large balls and generates coordinate information, the path planning algorithm calculates obstacle avoidance trajectories, the robotic arm positions itself according to the trajectory, and the crushing hand crushes the large balls. It is mainly used for automatically detecting and crushing large balls produced in a disc pelletizing machine, improving pelletizing quality and production efficiency.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of pelletizing equipment technology. More specifically, this invention relates to an automatic large pellet crushing device for a disc pelletizing machine. Background Technology

[0002] In industrial production, disc pelletizers are widely used in pelletizing processes, such as in mining and metallurgy, to turn powdered materials into spherical particles. However, during the disc pelletizing process, some large pellets with diameters exceeding the normal range often appear. The presence of these large pellets can have many adverse effects on pelletizing production.

[0003] Firstly, from the perspective of pelletizing quality, the presence of large pellets disrupts the homogeneity of the entire pelletizing system. Qualified pellets meet certain standards in terms of size and density. The inclusion of large pellets leads to an uneven size distribution, which in turn affects the roasting and sintering processes in subsequent steps. For example, during sintering, the larger size of the pellets results in slower internal heat and mass transfer, potentially leading to incomplete sintering and reduced strength and quality of the sinter.

[0004] Secondly, the presence of large balls can hinder the normal operation of the disc pelletizing machine. As the large balls roll within the pelletizing disc, the load on the disc increases, requiring the motors and other drive equipment to consume more energy to maintain its operation. Prolonged exposure to this high load will accelerate equipment wear, shorten its lifespan, and increase maintenance costs and downtime.

[0005] Furthermore, large balls can cause problems in subsequent conveying and processing stages. They may clog conveying pipes, chutes, and other equipment, affecting the normal transport of materials and causing production interruptions. Additionally, large balls may have difficulty passing through some screening equipment, requiring extra manual intervention, increasing labor costs and workload.

[0006] The causes of large pellets are complex. On the one hand, the properties of the material may fluctuate, such as uneven particle size distribution or unstable moisture content, which can lead to some material agglomerating and forming large pellets during the pelletizing process. On the other hand, unreasonable settings of the pelletizing machine's operating parameters, such as the rotational speed, tilt angle, water spray position, and water spray volume of the pelletizing disc, can also affect the pellet formation process and increase the probability of large pellets being produced.

[0007] In the past, the problem of large pellets was typically solved through manual screening and crushing. This method was not only inefficient and labor-intensive, but also difficult to process large pellets accurately and in real time. Furthermore, manual operation is inherently subjective and uncertain, potentially missing some large pellets or causing unnecessary damage to qualified pellets. With the continuous improvement of industrial automation, automated equipment has been explored to address the large pellet problem. However, current equipment on the market still falls short in terms of detection accuracy, crushing effect, and adaptability, failing to meet the needs of actual production. Therefore, developing a highly efficient and accurate automatic large pellet crushing device for disc pelletizers is of significant practical importance. Summary of the Invention

[0008] One object of the present invention is to solve at least the above-mentioned problems and to provide at least the advantages that will be described later.

[0009] To achieve these objectives and other advantages according to the present invention, an automatic large-ball crushing device for a disc pelletizer is provided, comprising: Image acquisition module, which is used to acquire image data of the ball-forming disc area; The control unit is communicatively connected to the image acquisition module and has a built-in machine vision-based image analysis algorithm and a three-dimensional path planning algorithm. The image analysis algorithm is configured to perform multi-target detection on the received image data, identify the outline of the sphere through an edge detection algorithm, calculate the equivalent diameter based on the projected area of ​​the sphere, and determine the presence of a large sphere when the detected equivalent diameter exceeds a preset threshold and the number of consecutive frames reaches a set value. The number of large spheres is counted, and when the number of large spheres reaches a preset threshold, three-dimensional coordinate information of the large sphere region is generated. The three-dimensional path planning algorithm is configured to calculate the obstacle avoidance trajectory based on the three-dimensional coordinate information of the large sphere region and the preset starting point position and broken area position. A six-degree-of-freedom robotic arm is positioned near the ball-forming plate area. The joint drive components of the six-degree-of-freedom robotic arm are signal-connected to the control unit. In response to the trigger command generated by the control unit, it performs spatial positioning based on the obstacle avoidance motion trajectory. The crusher includes a mounting base installed at the end of the six-degree-of-freedom robotic arm, multiple stationary fingers mounted on the mounting base, a drive assembly mounted on the mounting base, and multiple movable fingers driven to rotate by the drive assembly, wherein when the drive assembly drives the multiple movable fingers to rotate, the multiple movable fingers are inserted into the gaps between the multiple stationary fingers in a one-to-one correspondence.

[0010] Preferably, the static finger body is rod-shaped with a bent end, and has stop bars on both outermost sides.

[0011] Preferably, both sides of the stationary finger are provided with a comb-like structure, the roots of the comb-like structure are provided with triangular reinforcing ribs, and the gap G between the comb-like teeth is set as follows: G=d max +16mm, and G < D min , where d max For the diameter of a qualified sphere, D min The diameter of the large sphere; Both sides of the movable finger are provided with serrated structures corresponding to the comb-like structure. The tips of the serrated structures face the comb-like structure, and the heads of the serrated structures are double-sloping wedges with an included angle of 30°.

[0012] Preferably, the drive assembly includes a cylinder mounted on the mounting base, a connecting rod hinged to the piston rod of the cylinder, and a mounting rod hinged to the other end of the connecting rod, wherein the movable finger is fixed on the mounting rod, and the piston rod is in an extended state as the initial position, in which the movable finger and the stationary finger are in an open state.

[0013] Preferably, the device also includes a double eccentric wheel vibrator, which is mounted on the mounting base via a rubber spring composite damping layer. The stationary finger is mounted on the double eccentric wheel vibrator, wherein the vibration direction of the double eccentric wheel vibrator is set to vibrate obliquely at 45° relative to the length direction of the stationary finger. When the moisture content of the qualified balls is less than 10%, the vibration parameters of the double eccentric wheel vibrator are set as follows: frequency of 20Hz and amplitude of 2mm. When the moisture content of the qualified balls is not less than 10%, the vibration parameters of the double eccentric wheel vibrator are set as follows: frequency of 12Hz and amplitude of 5mm.

[0014] Preferably, the rubber spring composite damping layer includes a base plate, a telescopic sleeve disposed on the base plate, a spring sleeved inside the telescopic sleeve, and a top plate disposed on the top of the telescopic sleeve. The two ends of the spring are respectively fixed to the base plate and the top plate. The outer surface of the spring is covered with a rubber layer. The bottom surface and the top surface of the top plate are covered with a buffer layer. The double eccentric wheel vibrator is mounted on the top plate.

[0015] Preferably, the image analysis algorithm specifically includes the following steps: S1. Convert the received image data into a grayscale image and HSV color space. Based on the differences in the distribution of hue (H), saturation (S), and brightness (V) of the sphere and the material layer in the HSV color space, set a three-channel joint threshold range for dynamic segmentation and generate a preliminary binarization mask. S2. Extract the luminance channel of the HSV color space, calculate the mean luminance μ and variance σ² of the initially segmented region, and set the variance threshold σ. th For conditions satisfying σ²>σth The segmented region is determined to be a dynamic material layer, retaining σ²≤σ th The smooth regions are used as the candidate sphere set; S3. The image contrast (Contrast) is calculated using the gray-level co-occurrence matrix. The texture feature is used as the contrast, and a preset contrast threshold is determined based on the texture difference between the sphere and the material layer. C th The image is segmented to generate a binary mask for texture features, where the contrast is higher than a contrast threshold. C th The material layer area should not exceed the contrast threshold. C th As a candidate set of spheres; in, P ( i , j ) represents the pixel pairs in the gray-level co-occurrence matrix. i , j The joint probability of ) N The number of gray levels; S4. Fuse color and texture feature masks, and select pixels with gray values ​​in the range [μ-2σ, μ+2σ] and textures classified as spheres as initial seed points. Set a dynamic similarity criterion, iteratively merge adjacent similar pixels until convergence, and output the refined sphere region segmentation result. The pixel difference thresholds allowed by the dynamic similarity criterion are: gray value difference ΔGray≤15, HSV color distance ΔE≤8, and texture feature vector Euclidean distance ΔF≤0.2. S5. Use Canny edge detection to extract closed contours for the segmented spherical regions, calculate the convex hull and geometric parameters of the contours. The geometric parameters include area A, perimeter L, and shape factor SF = 4πA / L², where area A is the number of pixels. Set the geometric constraints as area threshold and shape factor SF ≥ 0.80. Contours that meet the geometric constraints are determined to be valid spheres. S6. Based on the relationship between the projected area of ​​the sphere and spatial geometry, establish an equivalent diameter model and calculate the equivalent diameter. D eq Trajectory association is performed on the detection results of multiple consecutive frames, and Kalman filtering is used to track the trajectory of the sphere. The number of stable trajectories is the total number of spheres in real time. in, A proj The projected area of ​​the sphere is calculated by multiplying the pixel area by the physical size of a single pixel. k θ is the ball disk tilt angle compensation coefficient, k = cosθ, and θ is the angle between the disk axis and the camera optical axis. S7. When the equivalent diameter is detected D eq Larger than the preset diameter threshold D th When the number of large balls exceeds a preset threshold, the coordinates of the large balls are reported and a breakage command is triggered.

[0016] The present invention has at least the following beneficial effects: First, this invention achieves automatic detection and crushing of large balls through the coordinated operation of an image acquisition module, a control unit, a six-degree-of-freedom robotic arm, and a crushing hand. The image analysis algorithm accurately identifies the large balls, the three-dimensional path planning algorithm ensures the robotic arm avoids obstacles to reach the ball's location, and the crushing hand effectively crushes the balls. This improves the automation and efficiency of ball production, reduces manual intervention, lowers labor intensity, and simultaneously improves ball quality.

[0017] Secondly, the design of the comb-like and serrated structures allows the moving and stationary fingers to more accurately crush large balls when working together. The reasonable spacing between the comb teeth ensures that qualified balls are not accidentally crushed, while also allowing large balls to be effectively crushed. Triangular reinforcing ribs enhance the structural strength of the stationary finger, improving its durability. The double-beveled wedge-shaped serrated head can cut into large balls more effectively, improving crushing efficiency.

[0018] Third, the drive assembly has a simple and reliable structure, enabling precise control of the moving fingers' movement. The extension and retraction of the cylinder's piston rod drives the connecting rod and mounting rod, thus achieving the opening and closing of the moving and stationary fingers. The initial position setting ensures that the moving and stationary fingers are open during the robotic arm's movement, avoiding interference with surrounding objects. The closing and crushing operation only occurs after reaching the large ball, improving operational accuracy and safety. Furthermore, the closing action of the fingers also helps to clean material adhering to the gaps between the fingers.

[0019] Fourth, to address the issue of qualified balls being easily broken, the double eccentric wheel vibrator operates before breaking large balls, using different vibration parameters to screen out qualified balls from the gap between the stationary fingers, leaving only large balls for breaking. This effectively avoids unnecessary damage to qualified balls, significantly improves the product qualification rate, and reduces resource waste.

[0020] Fifth, the image analysis algorithm, through multi-step processing, comprehensively utilizes color, texture, and geometric features to accurately identify large spheres from the image data of the sphere-forming plate area. The fusion of color and texture features and the application of dynamic similarity criteria improve the accuracy and refinement of sphere segmentation. Kalman filtering tracks the sphere's motion trajectory, making the detection of large spheres more stable and reliable. When conditions are met, the large sphere coordinates are reported and a crushing command is triggered, providing an accurate information basis for subsequent crushing operations, thereby improving the overall efficiency and accuracy of the automatic large sphere crushing device.

[0021] Sixth, the path planning algorithm accurately determines the workspace of the robotic arm and the position information of various objects by establishing a 3D map model. Calculating the optimal target point coordinates ensures that the robotic arm can efficiently grasp the large ball. The path search algorithm and optimization processing can find the optimal path from the starting point to each target point, and remove redundant nodes and smooth the path, making the movement of the robotic arm smoother and more efficient, while avoiding collisions with obstacles, thus improving the safety and reliability of the robotic arm operation.

[0022] Other advantages, objectives and features of the present invention will become apparent in part from the following description, and in part from those skilled in the art through study and practice of the invention. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the overall structure of the automatic crushing device according to one of the technical solutions of the present invention; Figure 2 This is a detailed view of the crushing hand according to one of the technical solutions of the present invention; Figure 3 This is a diagram showing two states of the crusher in one of the technical solutions of the present invention; Figure 4 This is a detailed view of the comb-like structure according to one of the technical solutions of the present invention; Figure 5 This is a detailed view of the triangular reinforcing rib in one of the technical solutions of the present invention; Figure 6 A detailed view of the serrated structure according to one of the technical solutions of the present invention; Figure 7 This is a detailed view of the rubber spring composite damping layer according to one of the technical solutions of the present invention.

[0024] Instruction manual diagram reference numerals: Image acquisition module 1, Six-DOF robotic arm 2, Ball-forming disc 3, Crushing hand 4, Mounting base 41, Stationary finger 42, Moving finger 43, Stop bar 44, Comb-like structure 5, Triangular reinforcing rib 6, Serrated structure 7, Cylinder 45, Connecting rod 46, Mounting rod 47, Double eccentric wheel vibrator 8, Rubber spring composite damping layer 9, Base plate 91, Telescopic sleeve 92, Spring 93, Top plate 94, Buffer layer 95, State when serrated structure is inserted between stationary fingers 10. Detailed Implementation

[0025] The present invention will now be described in further detail with reference to the accompanying drawings, so that those skilled in the art can implement it based on the description.

[0026] It should be noted that, unless otherwise specified, the experimental methods described in the following embodiments are all conventional methods, and the reagents and materials described are all commercially available unless otherwise specified. In the description of this invention, the orientation or positional relationship indicated by the terms is based on the orientation or positional relationship shown in the accompanying drawings, and is only for the convenience of describing this invention and simplifying the description. It does 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, and therefore should not be construed as a limitation of this invention.

[0027] like Figures 1-7 As shown, the present invention provides an automatic large ball crushing device for a disc ball pulverizer, comprising: Image acquisition module 1 is used to acquire image data of the ball-forming disk 3 area; The control unit is communicatively connected to the image acquisition module 1. It has a built-in machine vision-based image analysis algorithm and a three-dimensional path planning algorithm. The image analysis algorithm is configured to perform multi-target detection on the received image data, identify the outline of the sphere through the edge detection algorithm, calculate the equivalent diameter based on the projected area of ​​the sphere, and determine the presence of a large sphere when the detected equivalent diameter exceeds a preset threshold and the number of consecutive frames reaches a set value. The number of large spheres is counted, and when the number of large spheres reaches a preset threshold, the three-dimensional coordinate information of the large sphere area is generated. The three-dimensional path planning algorithm is configured to calculate the obstacle avoidance trajectory based on the three-dimensional coordinate information of the large sphere area and the preset starting point position and broken area position. A six-degree-of-freedom robotic arm 2 is positioned near the ball-forming disk 3. The joint drive components of the six-degree-of-freedom robotic arm 2 are signal-connected to the control unit. In response to the trigger command generated by the control unit, it performs spatial positioning based on the obstacle avoidance motion trajectory. The crusher 4 includes a mounting base 41 at the end of the six-degree-of-freedom robotic arm 2, multiple stationary fingers 42 mounted on the mounting base 41, a drive assembly mounted on the mounting base 41, and multiple movable fingers 43 driven to rotate by the drive assembly. When the drive assembly drives the multiple movable fingers 43 to rotate, each movable finger 43 corresponds to and inserts into the gaps between the multiple stationary fingers 42. After performing the crushing action, the movable fingers 43 continue to be driven to intersect the stationary fingers 42, effectively cleaning the material adhering to the corresponding gaps and achieving a maintenance-free operation.

[0028] In the above technical solution, addressing the issues of large balls affecting quality and operation in existing disc pelletizing machines, and the low efficiency and inaccuracy of manual processing, the image acquisition module 1 collects image data from the pelletizing disc 3 in real time. The image analysis algorithm of the control unit can accurately identify and count the large balls. When the number of large balls reaches a threshold, three-dimensional coordinate information is generated, allowing the robotic arm to automatically perform the crushing operation. This eliminates the inefficiency and inaccuracy of manual labor, achieving automatic detection and crushing of large balls, and improving the automation level and efficiency of pelletizing production.

[0029] Regarding the automatic detection and crushing of large balls, and improving automation and efficiency, the image acquisition module 1 provides the data foundation, the image analysis algorithm accurately identifies the large balls, the 3D path planning algorithm plans the obstacle avoidance trajectory, the six-degree-of-freedom robotic arm 2 accurately reaches the position of the large ball according to the trajectory, and the crushing hand 4 completes the crushing. The coordinated work of all parts automates the entire process, greatly improving production efficiency and reducing manual intervention and labor intensity.

[0030] To address the challenge of a robotic arm accurately reaching and safely returning to the position of a large sphere in a complex environment, the control unit's 3D path planning algorithm, based on the sphere's coordinates and preset position information, considers various objects and obstacles within the workspace to calculate the optimal obstacle-avoidance trajectory. The six-DOF robotic arm 2 exhibits high flexibility and positioning accuracy, enabling it to accurately move to the target position according to the trajectory and safely return to its initial position after completing the crushing operation.

[0031] Specifically, the implementation of image acquisition module 1: Image acquisition module 1 plays a crucial role in data collection within the entire large ball automatic crushing device. In practical applications, a high-definition industrial camera is recommended as the image acquisition device. The choice of installation location is critical; it must be installed at a suitable height and angle above the ball-forming plate 3 to ensure clear and comprehensive acquisition of image data of the ball-forming plate 3 area. Generally, the installation height should be determined based on the size of the ball-forming plate 3 and the camera's viewing angle. For example, for a ball-forming plate 3 with a diameter of 7.5 meters, the camera can be installed directly above the ball-forming plate 3 at a distance of 2-3 meters.

[0032] The camera should be installed vertically downwards to avoid blind spots. To ensure image quality, the camera should be cleaned and maintained regularly to prevent dust and moisture from affecting image clarity. For data transmission, a stable and reliable communication method, such as Ethernet or fiber optic communication, should be used to transmit the acquired image data to the control unit in real time and accurately.

[0033] Implementation of the control unit: The control unit is the core of the entire device, responsible for processing and analyzing image data and generating corresponding control commands. The control unit utilizes a high-performance industrial computer, possessing powerful computing and data processing capabilities, enabling it to rapidly process large amounts of image data.

[0034] The built-in machine vision-based image analysis algorithm is a crucial component of the control unit. During actual operation, the received image data is first preprocessed, including image enhancement and filtering, to improve image quality. Then, edge detection algorithms are used to identify the sphere's outline; commonly used edge detection algorithms include the Canny algorithm. The equivalent diameter is calculated based on the sphere's projected area, requiring the establishment of an accurate mathematical model of the projected area and equivalent diameter. When the detected equivalent diameter exceeds a preset threshold and the number of consecutive frames reaches a set value, a large sphere is identified, and its number is counted. When the number of large spheres reaches a preset threshold, the three-dimensional coordinate information of the large sphere region is generated.

[0035] The 3D path planning algorithm calculates the obstacle avoidance trajectory based on the 3D coordinates of the large sphere area and the preset starting point and broken area locations, using a path search algorithm such as the A* algorithm. During the calculation process, factors such as the robotic arm's workspace, the location and shape of obstacles must be fully considered to ensure the planned path is safe and efficient.

[0036] Implementation of the six-degree-of-freedom robotic arm 2: The six-degree-of-freedom robotic arm 2 is set up near the ball-forming plate 3. Its base needs to be firmly installed on the ground or other stable support structure to ensure the stability of the robotic arm during movement.

[0037] The joint drive components of the robotic arm are connected to the control unit via signals. When the control unit generates a trigger command, the joint drive components control the movement of each joint of the robotic arm according to the obstacle avoidance trajectory, achieving spatial positioning. During the movement of the robotic arm, its motion status is monitored in real time. Sensors installed at the joints acquire information such as joint angles and speeds, and this information is fed back to the control unit. The control unit adjusts the movement of the robotic arm based on the feedback information to ensure that the robotic arm can accurately reach the position of the large ball.

[0038] Meanwhile, to ensure the safety of the robotic arm, corresponding safety protection measures must be implemented, such as limit switches and overload protection devices. When the robotic arm moves beyond the preset range or encounters excessive resistance, the safety protection devices can be triggered in time to stop the robotic arm's movement, preventing equipment damage and personnel injury.

[0039] Implementation of the crusher 4: The crusher 4 is installed at the end of the six-degree-of-freedom robotic arm 2, and its mounting base 41 must be firmly fixed to the end of the robotic arm to ensure that there is no loosening or shaking during the crushing process.

[0040] The installation of the stationary finger 42 and the movable finger 43 must be precise to ensure that the movable finger 43 can accurately insert into the gap between the stationary fingers 42 during rotation. The drive assembly is the key component controlling the rotation of the movable finger 43, and typically uses a motor or cylinder 45 as the power source. In actual operation, when the robotic arm reaches the position of the large ball, the drive assembly drives the movable finger 43 to rotate according to the instructions of the control unit, inserting it into the gap between the stationary fingers 42 to crush the large ball. Figure 4 The diagram shows state 10 when the serrated structure is inserted between the still fingers.

[0041] To improve the crushing effect, the stationary finger 42 and the moving finger 43 should be made of materials with high strength and wear resistance, such as alloy steel. At the same time, the shape and size of the stationary finger 42 and the moving finger 43 should be optimized to better meet the crushing requirements of large balls.

[0042] In another technical solution, the static finger 42 is rod-shaped with a bent end, and has a stop bar 44 on both outermost sides.

[0043] In the above technical solution, to address the problem that large balls easily slip off the sides of the stationary finger 42 during crushing, leading to crushing failure and reduced efficiency, the design of the stationary finger 42 with its bent end and baffles 44 on both sides effectively prevents this. When the large ball enters the crushing area, the baffles 44 restrict its lateral movement, preventing it from escaping from the sides, thus ensuring that the large ball remains stably positioned between the stationary finger 42 and the moving finger 43. This improves the success rate of the crushing operation and guarantees crushing efficiency.

[0044] To address the issue of better limiting and fixing the large ball to improve the crushing effect, the stationary finger 42 is designed to enhance its ability to limit the large ball. The bent end initially blocks the rolling tendency of the large ball, while the side stops 44 further reinforce the constraint on the large ball, stabilizing its position during crushing. This allows the moving finger 43 to more accurately insert into the gap between the stationary finger 42 to crush the large ball, effectively improving the crushing effect.

[0045] Specifically, the main body of the stationary finger 42 adopts a rod-like structure: this structure has good mechanical properties and stability. Regarding material selection, considering the significant pressure it needs to withstand when crushing large balls, high-strength metallic materials, such as alloy steel, are typically chosen. Alloy steel has high strength and hardness, ensuring that the stationary finger 42 will not deform or be damaged during long-term use.

[0046] The dimensions of the rod-shaped structure should be determined based on the actual crushing requirements and the size of the large balls. For large balls to be crushed with a diameter of 45-300mm, the diameter of the rod is generally between 12-16mm, and the length is adjusted according to the overall design of the crusher 4 and the size of the ball-forming disc 3, usually between 200-350mm.

[0047] The bending of the end of the stationary finger 42 is implemented to better limit the movement of the large ball. The design of the bending angle is crucial and needs to be optimized based on the size of the large ball and the working method of the crusher 4. Generally, a bending angle between 30° and 60° is more suitable.

[0048] Installation of the stop lever 44: The stop lever 44 is installed on both outermost sides of the stationary finger 42 to prevent the large ball from sliding to the sides. The material of the stop lever 44 can also be the same high-strength metal material as the main body of the stationary finger 42 to ensure its strength and wear resistance.

[0049] Overall Assembly and Debugging: After machining and installing all parts of the stationary finger 42, it needs to be assembled onto the mounting base 41 of the crusher 4. During assembly, ensure the accurate installation position of the stationary finger 42 and the appropriate clearance between it and the moving finger 43. After installation, debugging is required to check whether the movement of the stationary finger 42 and the moving finger 43 is smooth and whether the stop bar 44 can effectively block the large ball. Simulated crushing experiments can be conducted to observe the movement of the large ball during the crushing process, allowing for adjustments and optimizations to the structure and installation of the stationary finger 42 to ensure it meets actual crushing requirements.

[0050] In another technical solution, both sides of the stationary finger 42 are provided with comb-like structures 5, the roots of the comb-like structures 5 are provided with triangular reinforcing ribs 6, and the gap G between the comb teeth is set as follows: G=d max +16mm, and G < D min , where d max For the diameter of a qualified sphere, D min The diameter of the large sphere; Both sides of the movable finger 43 are provided with serrated structures 7 corresponding to the comb-like structure 5. The tips of the serrated structures 7 face the comb-like structure 5, and the head of the serrated structure 7 is double-sloping wedge-shaped with an included angle of 30°.

[0051] In the above technical solution, the problem of accurately distinguishing between qualified balls and large balls and avoiding accidental breakage of qualified balls is cleverly solved by setting the comb tooth gap G. Since G=d max +16mm, and G < D min Qualified balls can pass smoothly through the gaps between the comb teeth, while large balls will be blocked, thus achieving accurate differentiation between qualified and large balls, effectively improving the product qualification rate and reducing resource waste.

[0052] To address the issue of insufficient structural strength in the stationary finger 42 and the moving finger 43, the triangular reinforcing ribs 6 at the root of the comb-like structure 5 play a crucial role. These triangular reinforcing ribs 6 enhance the structural strength of the stationary finger 42, making it less prone to deformation or damage when subjected to the pressure of the crushing ball, thus ensuring the stability of the crushing operation, extending the equipment's service life, and reducing maintenance costs.

[0053] Regarding the issue of low crushing efficiency of the moving finger 43, the serrated structure 7 on both sides of the moving finger 43 and the design of the head being a double-sloping wedge with an included angle of 30° greatly improve crushing efficiency. This structure can more effectively cut into the interior of large balls, crushing them quickly, reducing crushing time, and improving production efficiency.

[0054] Specifically, the comb-like structure 5 of the stationary finger 42 is implemented as follows: The comb-like structure 5 on both sides of the stationary finger 42 is a key component for achieving precise crushing. In terms of material selection, high hardness and wear resistance are required, and high-strength alloy steel is typically chosen. This material ensures that the comb teeth are not easily worn during long-term contact and compression with the pellets.

[0055] The dimensions of the comb teeth must be designed strictly according to the requirements of the comb tooth gap G. First, the diameter d of the qualified ball needs to be accurately measured. max and the diameter D of the large sphere min According to the formula G=d max +16mm, and G < D min Determine the gap between the comb teeth. The height of the comb teeth is generally between 5-10mm, and the width is between 5-15mm. The specific dimensions need to be adjusted according to the actual size of the pellets and the crushing requirements.

[0056] Implementation of triangular reinforcing rib 6: Triangular reinforcing rib 6 is placed at the root of the comb-like structure 5 to enhance the structural strength. The material of triangular reinforcing rib 6 is the same as the main material of the stationary finger 42 to ensure its connection strength with the comb teeth.

[0057] The dimensions of the triangular reinforcing rib 6 should be determined based on the dimensions of the comb teeth and the pressure it bears. Generally, the base length of the triangular reinforcing rib 6 is between 5-10mm, and the height is between 3-10mm. The triangular reinforcing rib 6 can be manufactured by welding or integral molding.

[0058] Implementation of the serrated structure 7 of the movable finger 43: The two side walls of the movable finger 43 are provided with serrated structures 7 corresponding to the comb-like structures 5. The material of the serrated structure 7 is also high-strength alloy steel to ensure its strength and wear resistance.

[0059] The tip of the serrated structure 7 faces the comb-shaped structure 5, which facilitates accurate cutting into the interior of the large ball when crushing it. The head of the serrated structure 7 is double-beveled wedge-shaped with an included angle of 30°. This angle design allows the moving finger 43 to insert into the large ball more smoothly, improving crushing efficiency.

[0060] The dimensions of the saw teeth should match the dimensions of the comb teeth. The height of the saw teeth is generally between 20-30mm, the upper diameter is 30-35mm, and the lower diameter is less than 5mm.

[0061] Overall assembly and debugging: After machining each part of the stationary finger 42 and the moving finger 43, they need to be assembled onto the mounting base 41 of the crusher 4. During the assembly process, it is necessary to ensure that the relative positions of the stationary finger 42 and the moving finger 43 are accurate and that the serrated structure 7 can be accurately inserted into the gap between the comb teeth.

[0062] After assembly, debugging is required. First, a no-load test is performed to check the smooth movement of the stationary finger 42 and the moving finger 43, and to check for any jamming. Then, a load test is performed by placing pellets of different sizes into the machine and observing whether qualified pellets can pass smoothly through the gaps between the comb teeth and whether large pellets can be effectively crushed. Based on the debugging results, the position and size of the stationary finger 42 and the moving finger 43 are fine-tuned to ensure the entire crushing device can operate normally and efficiently.

[0063] After a period of production statistics, the breakage rate of qualified balls was controlled within 5%, a reduction of 10% compared to before, while the breakage rate of large balls reached over 90%, an improvement of 5% compared to before. This indicates that the comb-like structure 5 of the stationary finger 42 and the serrated structure 7 of the moving finger 43 are reasonably designed, capable of accurately breaking large balls while protecting qualified balls to the greatest extent, thus improving the quality and efficiency of ball production.

[0064] In another technical solution, the drive assembly includes a cylinder 45 mounted on the mounting base 41, a connecting rod 46 hinged to the piston rod of the cylinder 45, and a mounting rod 47 hinged to the other end of the connecting rod 46. The movable finger 43 is fixed on the mounting rod 47. The piston rod is in an extended state, which is the initial position. In the initial position, the movable finger 43 and the stationary finger 42 are in an open state.

[0065] In the above technical solution, to address the issue of reliably driving the moving finger 43 to rotate and achieve accurate engagement with the stationary finger 42, the drive assembly, through a combination of cylinder 45, connecting rod 46, and mounting rod 47, can convert the linear motion of the piston rod of cylinder 45 into the rotation of the moving finger 43. This mechanical transmission method is stable and reliable, and can accurately control the rotation angle and speed of the moving finger 43, enabling precise engagement between the moving finger 43 and the stationary finger 42, effectively completing the large ball crushing operation, and improving the crushing success rate.

[0066] To prevent interference with surrounding objects during the movement of the robotic arm, the piston rod is set to the initial position with its extension extended, at which point the moving finger 43 and the stationary finger 42 are in an open state. This ensures that the moving finger 43 and the stationary finger 42 will not collide with or interfere with surrounding objects during the robotic arm's movement to the designated position, guaranteeing the safety of equipment operation and reducing the risk of equipment damage.

[0067] Regarding the issue of designing a simple and easy-to-maintain drive system to reduce costs, the drive assembly has a simple structure, consisting of common components such as cylinder 45, connecting rod 46, and mounting rod 47. The manufacturing process is relatively simple, reducing the equipment's manufacturing cost. At the same time, these components are easy to disassemble and replace, facilitating routine maintenance and repair, further reducing maintenance costs.

[0068] Specifically, the installation and adjustment of cylinder 45: Cylinder 45 is the power source for the drive assembly, and during installation, ensure it is securely fixed to the mounting base 41. Select a suitable installation position to ensure that the piston rod movement direction of cylinder 45 matches the rotation plane of the moving finger 43. During installation, use bolts to fix cylinder 45 to the designated position on the mounting base 41. When tightening the bolts, pay attention to even force to avoid tilting cylinder 45 during installation.

[0069] After installation, cylinder 45 needs to be tested. Connect the air supply and check the sealing of cylinder 45 to ensure there is no leakage. By controlling the switch of the air supply, test whether the extension and retraction of the piston rod of cylinder 45 is smooth and whether the stroke meets the design requirements. The stroke of the piston rod can be measured using tools such as vernier calipers for adjustment and calibration to ensure that the extension and retraction of the piston rod can accurately drive the movement of connecting rod 46.

[0070] Installation and adjustment of connecting rod 46: Connecting rod 46 transmits the movement of the piston rod of cylinder 45, and its two ends are hinged to the piston rod of cylinder 45 and mounting rod 47, respectively. When installing connecting rod 46, ensure the flexibility of the hinge joints to avoid jamming. Applying a suitable amount of lubricant to the hinge joints can reduce frictional resistance.

[0071] After installation, the length and angle of connecting rod 46 need to be adjusted. The length of connecting rod 46 must be precisely set according to the stroke of cylinder 45 and the rotation requirements of moving finger 43 to ensure that the extension and retraction of piston rod can accurately drive moving finger 43 to rotate. By adjusting the connection angle between connecting rod 46, piston rod, and mounting rod 47, it is ensured that moving finger 43 is in an open state with stationary finger 42 in the initial position, and can smoothly insert into the gap between stationary fingers 42 when piston rod retracts.

[0072] Installation and Fixing of Mounting Rod 47: Mounting rod 47 is used to fix the movable finger 43. During installation, ensure a secure connection between the mounting rod 47 and the movable finger 43. The movable finger 43 can be fixed to the mounting rod 47 by welding or bolting. Choose the appropriate connection method based on the structure and size of the movable finger 43. When welding, ensure welding quality to avoid problems such as incomplete welds or cracks. When using bolts, tighten the bolts to prevent the movable finger 43 from loosening.

[0073] The mounting rod 47 is hinged to the connecting rod 46. During installation, the concentricity of the hinge point must be ensured to allow the mounting rod 47 to rotate freely. Simultaneously, the position of the mounting rod 47 must be adjusted to ensure the precise engagement between the moving finger 43 and the stationary finger 42 during rotation. Fine-tuning the position and angle of the mounting rod 47 allows the moving finger 43 to accurately break the large ball when inserted into the gap between the stationary finger 42.

[0074] Overall debugging and optimization: After the installation of each component of the drive assembly is completed, overall debugging is required. First, perform no-load debugging, start cylinder 45, observe the rotation of moving finger 43, and check whether the opening and closing movements of moving finger 43 and stationary finger 42 are smooth, and whether there is any jamming or interference.

[0075] Then, load testing was performed by placing a simulated large ball to test the performance of the drive assembly in actual operation. The crushing effect of the moving finger 43 on the large ball was observed, and the drive assembly was optimized and adjusted based on the test results. For example, the air pressure of cylinder 45 was adjusted to change the movement speed and force of the piston rod; the position and angle of connecting rod 46 and mounting rod 47 were fine-tuned to improve the matching accuracy between the moving finger 43 and the stationary finger 42. After multiple debugging and optimizations, the drive assembly was ensured to work stably and reliably.

[0076] During actual operation, when the six-DOF robotic arm 2 moves to the designated position according to the control unit's instructions, the piston rod of cylinder 45 in the drive assembly is in an extended state, and the moving finger 43 and the stationary finger 42 are in an open state, preventing interference with surrounding equipment and pellets. When the robotic arm reaches the large ball position, the control unit issues a command, and the piston rod of cylinder 45 retracts. The retraction of the piston rod drives the mounting rod 47 to rotate via connecting rod 46, causing the moving finger 43 to rotate rapidly and insert into the gap between the stationary fingers 42, thus crushing the large ball. In 200 consecutive crushing operations, the accuracy rate of the moving finger 43 reached 99%, with no malfunctions or jamming. This fully demonstrates that the drive assembly has a simple and reliable structure, accurately controls the movement of the moving finger 43, effectively completes the large ball crushing task, and improves the efficiency and quality of pellet production.

[0077] In another technical solution, a double eccentric wheel vibrator 8 is also included, which is mounted on the mounting base 41 via a rubber spring composite damping layer 9. The stationary finger 42 is mounted on the double eccentric wheel vibrator 8, wherein the vibration direction of the double eccentric wheel vibrator 8 is set to vibrate obliquely at 45° relative to the length direction of the stationary finger 42. When the moisture content of the qualified balls is less than 10%, the vibration parameters of the double eccentric wheel vibrator 8 are set as follows: frequency of 20Hz and amplitude of 2mm. When the moisture content of the qualified balls is not less than 10%, the vibration parameters of the double eccentric wheel vibrator 8 are set as follows: frequency of 12Hz and amplitude of 5mm.

[0078] In the above technical solution, in order to address the problem that qualified balls are easily broken by accident, the double eccentric wheel vibrator 8 works first before breaking the large balls, and uses different vibration parameters to screen out qualified balls from the gap of the stationary finger 42, leaving only the large balls to be broken, which effectively avoids unnecessary damage to qualified balls, significantly improves the product qualification rate, and reduces resource waste.

[0079] Regarding the impact of vibration on equipment stability and lifespan, the rubber spring composite damping layer 9 provides excellent buffering and shock absorption. It absorbs and disperses the vibration energy generated by the double eccentric wheel vibrator 8, significantly reducing the vibration intensity transmitted to the mounting base 41 and other equipment components, ensuring stable equipment operation, extending equipment lifespan, and reducing maintenance costs.

[0080] Considering the impact of varying moisture content on the screening of qualified balls, adjusting vibration parameters according to moisture content is crucial. Using high-frequency, low-amplitude vibrations for low moisture content and low-frequency, high-amplitude vibrations for high moisture content better adapts to the characteristics of qualified balls with different moisture contents, improving screening efficiency and ensuring the reliability of the screening results.

[0081] Connection between stationary finger 42 and double eccentric wheel vibrator 8: The stationary finger 42 is installed on the double eccentric wheel vibrator 8. The connection method should ensure that the stationary finger 42 can fully receive the vibration transmitted by the vibrator.

[0082] Vibration parameter setting and adjustment: Accurately set the vibration parameters of the double eccentric wheel vibrator 8 according to the moisture content of the qualified balls. During production, the moisture content of the qualified balls should be checked regularly. When the moisture content of the qualified balls is less than 10%, set the frequency of the double eccentric wheel vibrator 8 to 20Hz and the amplitude to 2mm; when the moisture content of the qualified balls is not less than 10%, set the frequency to 12Hz and the amplitude to 5mm.

[0083] When setting vibration parameters, it can be done through the vibrator's control system. The control system generally has parameter display and adjustment functions, allowing operators to make corresponding settings based on the test results. After setting, a trial run should be conducted to observe the screening effect of qualified balls. If the screening effect is unsatisfactory, the vibration parameters should be further fine-tuned according to the actual situation until the optimal screening effect is achieved.

[0084] Control of the screening and crushing process: Before crushing large balls, the double eccentric wheel vibrator 8 is started to screen qualified balls according to the set vibration parameters. During the screening process, the flow of the balls and the screening effect should be observed. When most qualified balls have been screened out, the double eccentric wheel vibrator 8 is turned off. Then, the six-degree-of-freedom robotic arm 2 is controlled to move the crusher 4 to the crushing area. After reaching the crushing area, the drive assembly is activated, causing the moving finger 43 to insert into the gap between the stationary fingers 42 to crush the large balls. After crushing is completed, the robotic arm returns to the initial position, waiting for the next screening and crushing task.

[0085] Throughout the screening and crushing process, it is crucial to ensure the coordinated and accurate operation of all components. Automation can be achieved through control programming, improving both production efficiency and accuracy. Simultaneously, real-time monitoring of equipment operation is essential to promptly identify and address any anomalies.

[0086] During production, the moisture content of qualified balls is checked periodically. When the moisture content of qualified balls is found to be less than 10%, the frequency of the double eccentric wheel vibrator 8 is set to 20Hz and the amplitude to 2mm. After the vibrator is started, qualified balls are quickly screened out from the gap of the stationary fingers 42 under the action of vibration, while large balls remain on the stationary fingers 42. After a period of screening, the proportion of large balls mixed in with the screened qualified balls is extremely low, and the effective separation of qualified balls from large balls is basically achieved.

[0087] When the moisture content of the qualified balls increases to no less than 10%, the frequency of the vibrator is adjusted to 12Hz and the amplitude is adjusted to 5mm. At this time, although the flowability of the qualified balls deteriorates due to the increase in moisture, they can still be smoothly screened out through the 42 gap of the stationary fingers under the action of appropriate vibration parameters.

[0088] After screening, the double eccentric wheel vibrator 8 is turned off, and the robotic arm moves the crusher 4 to the crushing area. The drive assembly is activated, and the moving finger 43 accurately inserts into the gap between the stationary finger 42 to crush the large balls. After multiple tests, the crushing rate of the large balls reached over 95%, while the breakage rate of qualified balls was controlled within 0.5%.

[0089] In another technical solution, the rubber spring composite damping layer 9 includes a base plate 91, a telescopic sleeve 92 disposed on the base plate 91, a spring 93 sleeved inside the telescopic sleeve 92, and a top plate 94 disposed on the top of the telescopic sleeve 92. The two ends of the spring 93 are respectively fixed to the base plate 91 and the top plate 94. The outer surface of the spring 93 is covered with a rubber layer. The bottom surface and the top surface of the top plate 94 are covered with a buffer layer 95. The double eccentric wheel vibrator 8 is installed on the top plate 94.

[0090] In the above technical solution, the rubber spring composite damping layer 9 plays a significant role in addressing the problem of vibration damage to equipment. The spring 93 absorbs and buffers vibration energy, while the telescopic sleeve 92 ensures the stability of the spring 93 during its extension and retraction. The rubber layer further enhances the damping effect, dispersing and dissipating vibration energy. The buffer layers 95 above and below the top plate 94 also reduce impact force, effectively lowering the vibration intensity transmitted to the mounting base 41 and surrounding equipment, preventing equipment components from being damaged by vibration, and extending the equipment's service life.

[0091] Regarding the issue of noise generated by vibration, the rubber layer and buffer layer 95 have excellent sound insulation and absorption properties. They can reduce the transmission of structural noise caused by vibration, lower the noise level in the working environment, create a relatively quiet working environment for operators, and protect their health.

[0092] Regarding the issue of vibration affecting equipment stability and operational accuracy, this damping layer effectively reduces vibration, making the overall equipment more stable. Stable operation helps ensure the working accuracy of the automatic large ball crusher, improves crushing efficiency and product quality, and ensures continuous and stable equipment operation.

[0093] Installation of the rubber spring composite damping layer 9 and the double eccentric wheel vibrator 8: The assembled rubber spring composite damping layer 9 is installed on the mounting base 41 via the base plate 91 and secured with bolts, ensuring the tightening torque meets design requirements. Then, the double eccentric wheel vibrator 8 is installed on the top plate 94, also secured with bolts. During installation, ensure the double eccentric wheel vibrator 8 is positioned accurately, with a deviation not exceeding ±1mm. After installation, check the connection between the damping layer and the double eccentric wheel vibrator 8 for tightness and any looseness.

[0094] In another technical solution, the image analysis algorithm specifically includes the following steps: S1. Convert the received image data into a grayscale image and HSV color space. Based on the differences in the distribution of hue (H), saturation (S), and brightness (V) of the sphere and the material layer in the HSV color space, set a three-channel joint threshold range for dynamic segmentation and generate a preliminary binarization mask. S2. Extract the luminance channel of the HSV color space, calculate the mean luminance μ and variance σ² of the initially segmented region, and set the variance threshold σ. th For conditions satisfying σ²>σ th The segmented region is determined to be a dynamic material layer, retaining σ²≤σ th The smooth regions are used as the candidate sphere set; S3. The image contrast (Contrast) is calculated using the gray-level co-occurrence matrix. The texture feature is used as the contrast, and a preset contrast threshold is determined based on the texture difference between the sphere and the material layer. C th The image is segmented to generate a binary mask for texture features, where the contrast is higher than a contrast threshold. C th The material layer area should not exceed the contrast threshold. C th As a candidate set of spheres; in, P ( i , j ) represents the pixel pairs in the gray-level co-occurrence matrix. i , j The joint probability of ) N The number of gray levels; S4. Fuse color and texture feature masks, and select pixels with gray values ​​in the range [μ-2σ, μ+2σ] and textures classified as spheres as initial seed points. Set a dynamic similarity criterion, iteratively merge adjacent similar pixels until convergence, and output the refined sphere region segmentation result. The pixel difference thresholds allowed by the dynamic similarity criterion are: gray value difference ΔGray≤15, HSV color distance ΔE≤8, and texture feature vector Euclidean distance ΔF≤0.2. S5. Use Canny edge detection to extract closed contours for the segmented spherical regions, calculate the convex hull and geometric parameters of the contours. The geometric parameters include area A, perimeter L, and shape factor SF = 4πA / L², where area A is the number of pixels. Set the geometric constraints as area threshold and shape factor SF ≥ 0.80. Contours that meet the geometric constraints are determined to be valid spheres. S6. Based on the relationship between the projected area of ​​the sphere and spatial geometry, establish an equivalent diameter model and calculate the equivalent diameter. D eq Trajectory association is performed on the detection results of multiple consecutive frames, and Kalman filtering is used to track the trajectory of the sphere. The number of stable trajectories is the total number of spheres in real time. in, A proj The projected area of ​​the sphere is calculated by multiplying the pixel area by the physical size of a single pixel. k θ is the ball disk tilt angle compensation coefficient, k = cosθ, and θ is the angle between the disk axis and the camera optical axis. S7. When the equivalent diameter is detected D eq Larger than the preset diameter threshold D th When the number of large balls exceeds a preset threshold, the coordinates of the large balls are reported and a breakage command is triggered.

[0095] In the above technical solution, to address the problem of difficulty in accurately distinguishing between spheres and material layers, this image analysis algorithm converts image data to the HSV color space, uses the differences in hue, saturation, and brightness distribution for dynamic segmentation, and combines the gray-level co-occurrence matrix to calculate texture contrast. By integrating color and texture features, it can more accurately distinguish between spheres and material layers, improve the accuracy of large sphere recognition, and provide a reliable foundation for subsequent crushing operations.

[0096] To address the issue of insufficient refinement in spherical region segmentation, the algorithm integrates color and texture feature masks, selects pixels in specific grayscale ranges that are classified as spheres as initial seed points, and sets strict dynamic similarity criteria for iterative merging. This effectively reduces the impact of noise and interference, achieves refined segmentation of the spherical region, avoids over-segmentation and under-segmentation, and makes the determination of the number and position of large spheres more accurate.

[0097] Regarding the problem of tracking the trajectory of a sphere, the algorithm establishes an equivalent diameter model based on the sphere's projected area and spatial geometric relationship, and uses Kalman filtering to correlate and track the trajectory of multiple consecutive frames of detection results. This enables the algorithm to stably and accurately track the sphere's trajectory and obtain the quantity and position information of large spheres in real time, providing strong support for the real-time control of the automatic large sphere crushing device and improving the automation level and working efficiency of the device.

[0098] Specifically, step S1 is implemented as follows: In practical applications, the image data of the sphere-forming disk 3 region acquired by the image acquisition module 1 is first transmitted to the control unit. After receiving the image data, the control unit uses the corresponding image processing functions to convert the image data into a grayscale image and HSV color space. This conversion process can be achieved by calling functions in image processing libraries such as OpenCV.

[0099] After converting to the HSV color space, it is necessary to analyze the distribution differences of the sphere and the material layer in the three channels of hue (H), saturation (S), and brightness (V). This can be achieved by statistically analyzing a large number of image samples and examining the distribution range of pixel values ​​of the sphere and the material layer in different channels. For example, by analyzing 1000 images containing the sphere and the material layer, the pixel value range of the sphere and the material layer in the hue, saturation, and brightness channels can be determined.

[0100] Based on these analysis results, a reasonable three-channel joint threshold range was set. This threshold range required multiple trials and adjustments to ensure accurate differentiation between the sphere and the material layer. For example, experiments showed that setting the threshold range for hue (H) to [0,30], saturation (S) to [50,150], and brightness (V) to [80,200] achieved good initial segmentation of the sphere and the material layer. After setting the threshold range, dynamic segmentation was performed on the image to generate a preliminary binary mask.

[0101] Step S2: After generating the initial binarized mask, extract the luminance channel from the HSV color space. This can also be achieved using functions from an image processing library. After extracting the luminance channel, divide the initially segmented region into multiple sub-regions and calculate the mean luminance μ and variance σ² for each sub-region.

[0102] When calculating the mean brightness μ, the brightness values ​​of all pixels within the sub-region are summed and then divided by the total number of pixels in the sub-region. When calculating the variance σ², the difference between the brightness value of each pixel and the mean is squared according to the variance calculation formula, and then the average value is calculated.

[0103] Set the variance threshold σ th This threshold setting needs to be adjusted based on the actual situation. For example, through the analysis and experimentation of a large number of images, it was found that when the variance threshold σ...th Setting it to 50 allows for better differentiation between dynamic material layers and candidate sphere sets. For conditions satisfying σ²>σ... th The segmented region is determined to be a dynamic material layer, retaining σ²≤σ th The smooth regions are used as the candidate sphere set.

[0104] S3 Step Implementation: Image contrast (Contrast) is calculated using the Gray-Level Co-occurrence Matrix (GLCM). First, the calculation parameters of the GLCM are determined, such as the number of gray levels N and the offset. The number of gray levels N is generally determined based on the image's grayscale range and accuracy requirements, typically taking values ​​of 16, 32, or 64. The offset can be chosen in different directions and distances depending on the specific requirements, such as horizontal, vertical, or diagonal.

[0105] When calculating the gray-level co-occurrence matrix, the frequency of occurrence of pixel pairs (i,j) satisfying a specific offset in the image is statistically analyzed to obtain the gray-level co-occurrence matrix P(i,j). Then, the contrast ratio Contrast of the image is calculated according to the formula for contrast ratio.

[0106] The contrast threshold C is preset based on the texture difference between the sphere and the material layer. th The setting of this threshold also needs to be determined through the analysis and experimentation of a large number of images. For example, experiments have shown that when the contrast threshold C... th When set to 0.3, the material layer region and the candidate sphere set can be distinguished relatively well. The image is segmented to generate a binary texture feature mask, where the contrast is higher than the contrast threshold C. th The material layer area should not exceed the contrast threshold C. th This serves as a candidate set of spheres.

[0107] Step S4: Merge the color and texture feature masks, that is, merge the initial binary mask and the texture feature binary mask. This merging can be achieved through logical operations, such as AND and OR operations.

[0108] Pixels with grayscale values ​​in the range [μ-2σ, μ+2σ] and classified as spheres are selected as initial seed points. This range is set based on statistical principles, which can effectively select pixels with features similar to those of a sphere. Dynamic similarity criteria are set, including grayscale difference ΔGray ≤ 15, HSV color distance ΔE ≤ 8, and Euclidean distance of texture feature vectors ΔF ≤ 0.2.

[0109] Based on a dynamic similarity criterion, adjacent similar pixels are iteratively merged. In each iteration, the similarity between an adjacent pixel and the initial seed point is checked. If the similarity criterion is met, the pixel is merged into the spherical region. This process is repeated until no new pixels can be merged, and the refined spherical region segmentation result is output.

[0110] S5 Step Implementation: The segmented spherical region is processed using Canny edge detection to extract closed contours. The Canny edge detection algorithm is a classic edge detection algorithm that can accurately extract edge information in an image through steps such as Gaussian filtering, calculating gradient magnitude and direction, non-maximum suppression, and double threshold detection.

[0111] After extracting the closed contour, calculate the convex hull and geometric parameters of the contour. When calculating the area A, count the number of pixels within the contour; when calculating the perimeter L, traverse the pixels on the contour, calculate the distance between adjacent pixels and sum them; calculate the shape factor SF = 4πA / L².

[0112] Set the geometric constraints to an area threshold and a shape factor SF ≥ 0.80. The area threshold needs to be adjusted according to the actual situation. For example, setting the area threshold to 100 pixels can effectively filter out spheres that meet the requirements. Contours that satisfy the geometric constraints are considered valid spheres.

[0113] Step S6 Implementation: Based on the relationship between the projected area of ​​the sphere and spatial geometry, establish an equivalent diameter model. First, calculate the projected area Aproj of the sphere region, obtained by converting pixel area × single pixel physical size. The single pixel physical size can be determined through camera calibration.

[0114] Determine the tilt compensation coefficient k for the sphere, k = cosθ, where θ is the angle between the sphere's axis and the camera's optical axis. This angle can be obtained through measurement or pre-calibration. Calculate the equivalent diameter Deq based on the equivalent diameter model.

[0115] Trajectory association is performed on the detection results of multiple consecutive frames. Feature-based matching methods, such as matching based on shape features, color features, etc., can be used to associate spheres in different frames. Kalman filtering is used to track the sphere's motion trajectory. Kalman filtering is a commonly used state estimation method that can effectively track the sphere's motion trajectory through prediction and update steps. The number of stable trajectories is the total number of spheres in real time.

[0116] Step S7 Implementation: Set the preset diameter threshold D th And preset quantity thresholds. These thresholds need to be determined based on actual production needs and the definition of a large ball. For example, when the preset diameter threshold D... thWhen the setting is 50mm and the preset quantity threshold is set to 5, it can meet the needs of large ball detection in production.

[0117] When the detected equivalent diameter Deq is greater than the preset diameter threshold D th When the number of large balls exceeds a preset threshold, the coordinates of the large balls are reported and a crushing command is triggered. The control unit sends the coordinate information of the large balls to the joint drive assembly of the six-degree-of-freedom robotic arm 2. The robotic arm moves to the designated position according to the coordinate information and drives the crusher 4 to crush the large balls.

[0118] After a period of operation, the image analysis algorithm can accurately identify large balls, with an accuracy rate of over 99%, which is 10% higher than before. This effectively improves the working efficiency and accuracy of the automatic large ball crushing device, reduces manual intervention, and lowers labor intensity.

[0119] In another technical solution, the path planning algorithm specifically includes the following steps: Step 1: Using the ball-making plate 3 as the center, determine the working space of the robotic arm, establish a three-dimensional map model, and mark the position and shape information of the ball-making plate 3, the feeding chute, the robotic arm base, and obstacles on the map. Step 2: Using the initial position of the robotic arm as the starting point, calculate the optimal target point coordinates for the robotic arm to grasp the large ball based on the position of the large ball; Step 3: Employ a path search algorithm to search the constructed map for the optimal path from the starting point to the target point for grabbing the large ball, then to the target point above the unloading conveyor, and finally back to the starting point. Step 4: Optimize the found optimal path by removing redundant nodes and smoothing the path. Then, send the optimized and smoothed path information to the joint driver component.

[0120] In the above technical solution, to address the issues of determining the working space of the robotic arm and avoiding collisions, a three-dimensional map model is established with the ball-making disk 3 as the center, and the position and shape information of each object are marked. This allows the robotic arm to clearly understand the working environment, plan its movement path in advance, effectively avoid collisions with surrounding objects, ensure the safe operation of the equipment and the safety of personnel, and also reduce the damage and maintenance costs caused by collisions.

[0121] To calculate the optimal target point coordinates for a robotic arm to grasp a large ball, the initial position of the robotic arm is used as the starting point, and the calculation is performed in conjunction with the position of the large ball. This ensures that the robotic arm grasps the large ball in the optimal posture and position, improving the efficiency and accuracy of grasping, reducing ineffective movements, saving time and energy, and making the large ball crushing operation more efficient.

[0122] Regarding the issue of finding the optimal path to reduce motion time and energy consumption, a path search algorithm is used to search the map for a path from the starting point to each target point and back to the starting point. This can find a relatively shortest and most reasonable path, allowing the robotic arm to move along this path, which greatly reduces motion time and energy consumption, improves work efficiency, and also reduces the operating cost of the equipment.

[0123] Traditional pathfinding algorithms involve complex search and evaluation processes, such as traversing and calculating numerous unnecessary nodes. The simplified pathfinding algorithm in this paper, through in-depth analysis of the robotic arm's workspace and production scenario, pre-sets starting positions, etc. For example, based on the fixed positions and shapes of the ball-forming tray 3, the unloading chute, and obstacles, the feasible and prohibited areas for the robotic arm's movement are determined. During path search, nodes clearly outside the feasible area are directly excluded, reducing the search range and thus avoiding calculations for these nodes. Furthermore, for some path segments that can be determined through simple geometric relationships or empirical formulas, complex algorithmic calculations are no longer performed; instead, a pre-determined optimal solution is directly adopted, further reducing the computational load.

[0124] To address the issue of redundant nodes in the path causing uneven motion, the optimal path is optimized by removing redundant nodes and smoothing the path, making the robotic arm's motion path simpler and smoother. This improves the robotic arm's motion performance, reduces vibration and impact during movement, and extends the robotic arm's lifespan.

[0125] Specifically, step one is implemented as follows: Determine the working space of the robotic arm: First, use measuring tools, such as a laser rangefinder and a total station, to accurately measure the pelletizing tray 3, the unloading chute, the robotic arm base, and any potential obstacles. Measurements should include the dimensions (length, width, height) and relative positions (distance from the center of the pelletizing tray 3, angles, etc.) of each object.

[0126] For a robotic arm, its maximum extension range and minimum retraction range in space are determined by its kinematic model and actual range of motion, thereby defining the workspace of the robotic arm. For example, each joint of a six-DOF robotic arm 2 has a certain range of rotation angles. By calculating the combination of these angle ranges in space, the reachable space of the robotic arm is obtained.

[0127] Establishing a 3D map model: Using professional 3D modeling software such as AutoCAD and SolidWorks, a 3D map model is established with the ball-forming plate 3 as the center, based on the measured data. During the modeling process, the 3D models of the ball-forming plate 3, the feeding sluice, the robotic arm base, and obstacles are accurately drawn according to their actual dimensions and positional relationships. To facilitate subsequent path planning, each object is numbered and labeled, recording its position and shape information. For example, the diameter and height of the ball-forming plate 3, and the tilt angle and length of the feeding sluice are labeled.

[0128] Labeling Location and Shape Information: In the 3D map model, different colors, lines, or symbols are used to distinguish different objects, and their location and shape information are labeled in detail. For example, red lines represent obstacles, and blue lines represent the outline of the ball-building disk 3. At the same time, the specific size and location coordinates of each object are recorded next to the model or in a separate document.

[0129] Step Two Implementation: Obtaining the initial position of the robotic arm: In the control system of the robotic arm, sensors such as encoders installed at the joints acquire the angle information of each joint in real time. Based on the kinematic model of the robotic arm, this joint angle information is converted into the coordinates of the robotic arm's end effector in three-dimensional space, thereby determining the initial position of the robotic arm, which is used as the starting point.

[0130] Obtaining the position of the large ball: Image acquisition module 1 acquires image data of the area of ​​the ball-forming disk 3. After processing by the image analysis algorithm in the control unit, the position information of the large ball in the image is obtained. Then, according to the camera calibration parameters, the position information in the image is converted into three-dimensional coordinates in actual space to determine the position of the large ball.

[0131] Calculating the optimal target point coordinates: Based on the structure and motion characteristics of the robotic arm, and the position of the large ball, a mathematical algorithm is used to calculate the optimal target point coordinates for the robotic arm to grasp the large ball. For example, an inverse kinematics algorithm can be used, combined with the kinematic model of the robotic arm, to calculate the optimal posture and coordinates of the robotic arm's end effector reaching the position of the large ball under the grasping conditions. During the calculation process, factors such as the joint limitations and range of motion of the robotic arm must be considered to ensure that the calculated target point coordinates are feasible.

[0132] Step 3 implementation: Choosing a pathfinding algorithm: Common pathfinding algorithms include A* algorithm and Dijkstra's algorithm. Select a suitable algorithm based on the specific situation. For example, the A* algorithm combines the advantages of heuristic search and graph search, and can find the optimal path from the starting point to the target point in a relatively short time.

[0133] Path searching on the map: The starting point, the target point for grabbing the large ball, the target point above the discharge sluice, and the starting point (return point) are input into the path search algorithm. The algorithm searches within the constructed 3D map model, gradually finding the path from the starting point to the target point for grabbing the large ball, then to the target point above the discharge sluice, and finally back to the starting point by continuously expanding the nodes and calculating the distances and heuristic function values ​​between nodes.

[0134] During the search process, the algorithm avoids areas marked as obstacles to ensure that the searched path is feasible. Simultaneously, it records the coordinates of each node along the searched path.

[0135] Step Four: Redundant node removal: Analyze the nodes on the searched path to determine which nodes are redundant. For example, if the path between two nodes is a straight line without obstacles, then the intermediate node between these two nodes may be redundant. By calculating the distance and direction between nodes, these redundant nodes are removed, making the path simpler.

[0136] Smoothing: The path is smoothed using methods such as curve fitting. For example, spline curves can be used to fit nodes on the path, making the path smoother and reducing abrupt acceleration changes during the robot arm's movement. During the smoothing process, it is essential to ensure that the path still meets the robot arm's kinematic constraints and workspace limitations.

[0137] Sending Path Information: The optimized and smoothed path information is processed and converted into a format recognizable by the joint drive component, and then sent to the joint drive component. Based on the received path information, the joint drive component controls the movement of each joint of the robotic arm, causing the robotic arm to operate along the planned path.

[0138] One example: First, using tools such as a laser rangefinder and a total station, precise measurements were taken of the pelletizing disc 3 (6 meters in diameter and 1.5 meters in height), the unloading chute (30-degree inclination angle and 4 meters in length), the base of the six-degree-of-freedom robotic arm 2, and the surrounding obstacles. Then, using AutoCAD software, a three-dimensional map model was created with the pelletizing disc 3 as the center, and the position and shape information of each object were marked in detail.

[0139] When the image acquisition module 1 detects a large ball in the ball-forming tray 3, the control unit obtains the initial position coordinates (1,1,1) (in meters) of the robotic arm, and simultaneously obtains the position coordinates (3,3,0.5) of the large ball based on the image analysis algorithm. The optimal target point coordinates (3,3,0.6) for the robotic arm to grasp the large ball are calculated using the inverse kinematics algorithm to ensure that the robotic arm can stably grasp the large ball.

[0140] Next, the A* algorithm is used to search for a path in the 3D map model, from the starting point (1,1,1) to the target point (3,3,0.6) for grabbing the large ball, then to the target point (4,2,1) above the discharge chute, and finally back to the starting point (1,1,1). The algorithm successfully avoids the obstacles marked on the map and finds a feasible path.

[0141] Finally, the searched path was optimized by removing five redundant nodes and smoothing it using spline curves. The optimized and smoothed path information was sent to the joint drive component, and the robotic arm accurately moved to the target position according to the planned path, successfully grasped and crushed the large ball, and then safely returned to the starting point. After multiple tests, the path planning algorithm can effectively plan the movement path of the robotic arm, improving the efficiency and safety of large ball crushing.

[0142] Although embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the specification and embodiments. They can be applied to various fields suitable for the present invention. For those skilled in the art, other modifications can be easily made. Therefore, without departing from the general concept defined by the claims and their equivalents, the present invention is not limited to the specific details and illustrations shown and described herein.

Claims

1. An automatic large-ball crushing device for a disc pelletizing machine, characterized in that, include: Image acquisition module, which is used to acquire image data of the ball-forming disc area; The control unit is communicatively connected to the image acquisition module and has a built-in machine vision-based image analysis algorithm and a three-dimensional path planning algorithm. The image analysis algorithm is configured to perform multi-target detection on the received image data, identify the outline of the sphere through an edge detection algorithm, calculate the equivalent diameter based on the projected area of ​​the sphere, and determine the presence of a large sphere when the detected equivalent diameter exceeds a preset threshold and the number of consecutive frames reaches a set value. The number of large spheres is counted, and when the number of large spheres reaches a preset threshold, three-dimensional coordinate information of the large sphere region is generated. The three-dimensional path planning algorithm is configured to calculate the obstacle avoidance trajectory based on the three-dimensional coordinate information of the large sphere region and the preset starting point position and broken area position. A six-degree-of-freedom robotic arm is positioned near the ball-forming plate area. The joint drive components of the six-degree-of-freedom robotic arm are signal-connected to the control unit. In response to the trigger command generated by the control unit, it performs spatial positioning based on the obstacle avoidance motion trajectory. The crusher includes a mounting base installed at the end of the six-degree-of-freedom robotic arm, multiple stationary fingers mounted on the mounting base, a drive assembly mounted on the mounting base, and multiple movable fingers driven to rotate by the drive assembly, wherein when the drive assembly drives the multiple movable fingers to rotate, the multiple movable fingers are inserted into the gaps between the multiple stationary fingers in a one-to-one correspondence. Both sides of the stationary finger are provided with a comb-like structure, and the roots of the comb-like structure are provided with triangular reinforcing ribs. The gap G between the comb-like teeth is set as follows: G=d max +16mm, and G < D min , where d max For the diameter of a qualified sphere, D min The diameter of the large sphere; Both sides of the movable finger are provided with serrated structures corresponding to the comb-like structure. The tips of the serrated structures face the comb-like structure. The head of the serrated structure is double-sloping wedge-shaped with an included angle of 30°. It also includes a double eccentric wheel vibrator, which is mounted on the mounting base via a rubber spring composite damping layer. The stationary finger is mounted on the double eccentric wheel vibrator, wherein the vibration direction of the double eccentric wheel vibrator is set to vibrate obliquely at 45° relative to the length direction of the stationary finger. When the moisture content of the qualified balls is less than 10%, the vibration parameters of the double eccentric wheel vibrator are set as follows: frequency of 20Hz and amplitude of 2mm. When the moisture content of the qualified balls is not less than 10%, the vibration parameters of the double eccentric wheel vibrator are set as follows: frequency of 12Hz and amplitude of 5mm.

2. The automatic large ball crushing device for a disc pelletizer as described in claim 1, characterized in that, The static finger body is rod-shaped with a bent end, and has stop bars on both outermost sides.

3. The automatic large ball crushing device for a disc pelletizer as described in claim 1, characterized in that, The drive assembly includes a cylinder mounted on the mounting base, a connecting rod hinged to the piston rod of the cylinder, and a mounting rod hinged to the other end of the connecting rod. The movable finger is fixed on the mounting rod. The piston rod is in an extended state, which is the initial position. In the initial position, the movable finger and the stationary finger are in an open state.

4. The automatic large ball crushing device for a disc pelletizer as described in claim 1, characterized in that, The rubber spring composite damping layer includes a base plate, a telescopic sleeve disposed on the base plate, a spring sleeved inside the telescopic sleeve, and a top plate disposed on the top of the telescopic sleeve. The two ends of the spring are respectively fixed to the base plate and the top plate. The outer surface of the spring is covered with a rubber layer. The bottom surface and the top surface of the top plate are covered with a buffer layer. The double eccentric wheel vibrator is installed on the top plate.

5. The automatic large ball crushing device for a disc pelletizer as described in claim 1, characterized in that, The image analysis algorithm specifically includes the following steps: S1. Convert the received image data into a grayscale image and HSV color space. Based on the differences in the distribution of hue (H), saturation (S), and brightness (V) of the sphere and the material layer in the HSV color space, set a three-channel joint threshold range for dynamic segmentation and generate a preliminary binarization mask. S2. Extract the luminance channel of the HSV color space, calculate the mean luminance μ and variance σ² of the initially segmented region, and set the variance threshold σ. th For conditions satisfying σ²>σ th The segmented region is determined to be a dynamic material layer, retaining σ²≤σ th The smooth regions are used as the candidate sphere set; S3. The image contrast (Contrast) is calculated using the gray-level co-occurrence matrix. The texture feature is used as the contrast, and a preset contrast threshold is set based on the texture difference between the sphere and the material layer. C th The image is segmented to generate a binary mask for texture features, where the contrast is higher than a contrast threshold. C th The material layer area should not exceed the contrast threshold. C th As a candidate set of spheres; in, P ( i , j ) represents the pixel pairs in the gray-level co-occurrence matrix. i , j The joint probability of ) N The number of gray levels; S4. Fuse color and texture feature masks, and select pixels with gray values ​​in the range [μ-2σ, μ+2σ] and textures classified as spheres as initial seed points. Set a dynamic similarity criterion, iteratively merge adjacent similar pixels until convergence, and output the refined sphere region segmentation result. The pixel difference thresholds allowed by the dynamic similarity criterion are: gray value difference ΔGray≤15, HSV color distance ΔE≤8, and texture feature vector Euclidean distance ΔF≤0.

2. S5. Use Canny edge detection to extract closed contours for the segmented spherical regions, calculate the convex hull and geometric parameters of the contours. The geometric parameters include area A, perimeter L, and shape factor SF = 4πA / L², where area A is the number of pixels. Set the geometric constraints as area threshold and shape factor SF ≥ 0.

80. Contours that meet the geometric constraints are determined to be valid spheres. S6. Based on the relationship between the projected area of ​​the sphere and spatial geometry, establish an equivalent diameter model and calculate the equivalent diameter. D eq Trajectory association is performed on the detection results of multiple consecutive frames, and Kalman filtering is used to track the trajectory of the sphere. The number of stable trajectories is the total number of spheres in real time. in, A proj The projected area of ​​the sphere is calculated by multiplying the pixel area by the physical size of a single pixel. k θ is the ball disk tilt angle compensation coefficient, k = cosθ, and θ is the angle between the disk axis and the camera optical axis. S7. When the equivalent diameter is detected D eq Larger than the preset diameter threshold D th When the number of large balls exceeds a preset threshold, the coordinates of the large balls are reported and a break command is triggered.

Citation Information

Patent Citations

  • Disc pelletizer and method for crushing large pellets in disc pelletizer

    CN114247379A

  • Device for picking up large balls by robot in iron ore pellet preparation process

    CN215240883U