Robot disc type conveyor belt following method
By refreshing the workpiece pose in the reference coordinate system, planning accelerated motion, and switching to the workpiece coordinate system to maintain synchronization, combined with trajectory planning and overlay technology, the problem of lack of model and error in the robot disc conveyor follower scheme is solved, and high-precision follower control is achieved.
Patent Information
- Application Number
- CN202610177597.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-02-07
- Publication Date
- 2026-05-15
AI Technical Summary
Existing robot disc conveyor belt following schemes lack clear following models and implementation methods, and the position approximation method and position prediction method have uncertain following errors, making it difficult to guarantee following accuracy.
By refreshing the workpiece pose in the reference coordinate system, planning accelerated motion to keep the robot and workpiece relatively stationary, switching to the workpiece coordinate system to maintain synchronization, and combining trajectory planning and trajectory overlay technology, the pose synchronization of key points of the robot and workpiece is achieved.
It achieves a clear following model, avoids uncertain errors, ensures following accuracy and smooth robot motion, and is suitable for industrial mass production.
Smart Images
Figure CN122033943A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of robot control technology, and in particular to a method for robot following a disc-shaped conveyor belt. Background Technology
[0002] Robotic conveyor belt following is a crucial technology in industrial automation. It uses sensors to detect moving objects on a conveyor belt and controls a robot to follow those objects along a trajectory, enabling operations such as grasping, processing, and inspecting workpieces on the conveyor belt. Based on the type of conveyor belt, robotic conveyor belt following is mainly divided into two categories: linear conveyor belt following and disc-type conveyor belt following. On a linear conveyor belt, the object's movement path is a straight line, and its posture does not change with the conveyor belt's movement, making following control relatively simple. However, on a disc-type conveyor belt, the object's movement path is an arc, and its posture changes synchronously with the conveyor belt's movement, significantly increasing the difficulty of robot following control on disc-type conveyor belts.
[0003] In existing technologies, the implementation steps of robot disc-type conveyor belt following typically include a detection step and an execution step. For the detection step, firstly, a task queue is established; then, workpieces are identified and their coordinates are obtained, while the rotation information of the disc is acquired via an encoder. The workpiece coordinates and the current disc rotation information are stored in the task queue as feature information for the corresponding preset task; finally, the task queue is traversed to determine if each preset task overlaps with a following task in the queue, discarding the preset task or inserting it into the task queue. For the execution step, firstly, a work area is established on the disc; then, following tasks are retrieved sequentially from the task queue, and it is determined whether the following task can be completed. If it can, the following task is executed; otherwise, it is abandoned. Finally, a cleanup step is performed based on the contents of the task queue.
[0004] Currently, there are two main methods for implementing robot-driven disc conveyor belt following: position approximation and position prediction. Position approximation uses the positional deviation between the robot and the workpiece, along with the rotation information of the disc, to control the robot to approximate and follow the workpiece. Position prediction, similar to a motion encounter problem, uses multiple tests to obtain the average motion time of the robot between relevant path points. It then uses this average motion time and the disc's rotation information to predict the workpiece's pose, thus enabling the robot to follow the workpiece.
[0005] There are several specific implementation schemes in the existing technology, but all of them have obvious defects: 1. Patent CN113232022A discloses a disk conveying tracking control method, system, device, and storage medium. This scheme establishes a moving worktable coordinate system on the disk based on a standby point, midpoint, and end point; acquires images of materials in the work area and corresponding target positions based on a preset frequency; acquires disk motion information output by the encoder on the disk; creates preset tasks for each identified object; and uses the target position of the object and the current disk motion information as feature information for the corresponding preset task. However, this scheme does not describe a clear following method, making it impossible to effectively achieve dynamic tracking of the workpiece by the robot in practical applications.
[0006] 2. Patent CN114332164A discloses a semiconductor workpiece gripping method and control system using a disk conveyor. This method determines the first coordinate of the workpiece in a reference coordinate system; uses this first coordinate as a following interpolation position to control the robot to move to the first coordinate; and controls the robot to approximate and follow the workpiece based on the following interpolation position, the first coordinate, and the rotation information of the disk. However, the position approximation following method used in this method is prone to uncertain following errors due to the circular motion characteristics of the disk conveyor belt and the dynamic changes in the workpiece's posture, making it difficult to guarantee good following accuracy.
[0007] 3. The journal article "A Picking Strategy for Circular Conveyor Tracking" describes the coordinates of a workpiece on a disk using homogeneous transformation and employs a prediction method to obtain the workpiece's pose. The key prediction parameter is the robot's average motion time, which is obtained through multiple motion tests between relevant points. However, this position prediction tracking method is also prone to uncertain tracking errors due to the average motion time being affected by factors such as robot load and motion path, making it difficult to guarantee stable tracking accuracy.
[0008] In summary, existing robotic disc conveyor belt following solutions either lack a clear following model and implementation method or have uncertain following errors, failing to meet the following accuracy requirements in industrial production. Therefore, a technical solution that can solve the above problems is urgently needed. Summary of the Invention
[0009] The purpose of this invention is to overcome the problems of the prior art and provide a robot disc conveyor belt following method to solve the technical problems that existing robot disc conveyor belt following schemes lack clear and complete following models and implementation methods, and that position approximation methods and prediction methods have uncertain following errors, making it difficult to guarantee following accuracy.
[0010] The above objectives are achieved through the following technical solutions: A method for robot following a disc-shaped conveyor belt includes the following steps: Step (1): Move the robot to the current pose of the workpiece. Refresh the workpiece pose, obtain the robot pose, convert the workpiece pose and robot pose to a reference coordinate system, determine the pose deviation between the robot pose and the workpiece pose in the reference coordinate system, perform trajectory planning on the pose deviation, and control the robot to move to the current pose of the workpiece in the reference coordinate system. Step (2): Keep the robot and the workpiece relatively stationary. Plan an acceleration motion and control the robot to move along the arc-shaped movement path of the workpiece. When the acceleration motion ends, the robot's linear velocity and angular velocity in the reference coordinate system are consistent with the workpiece, and the robot and the workpiece reach a state of relative stillness. Step (3): Synchronize the poses of the key points on the robot and the workpiece. First, switch the robot's reference coordinate system to the workpiece coordinate system to maintain a relatively static pose constraint relationship between the robot and the workpiece. Then, determine the pose deviation between the robot's pose and the pose of key points on the workpiece in the workpiece coordinate system, and perform trajectory planning on the pose deviation. Finally, based on the pose constraint of the robot and the workpiece being relatively static, superimpose the deviation trajectory of the robot's pose and the pose of key points on the workpiece, so that the robot moves to the pose of the key points on the workpiece, and finally maintains a pose synchronization state with the key points on the workpiece.
[0011] Furthermore, the initial value of the workpiece pose in step (1) is obtained by real-time detection of the workpiece on the disc-shaped conveyor belt by a vision sensor. The vision sensor collects image information of the workpiece at a preset frequency and extracts the pose data of the workpiece based on an image recognition algorithm.
[0012] Furthermore, when planning an acceleration motion in step (2), the magnitude of the acceleration and the acceleration time of the acceleration motion are determined based on the curvature of the workpiece's arc-shaped moving path, the current linear velocity of the workpiece, and the rotational angular velocity, so that the robot can smoothly reach a motion state consistent with the workpiece.
[0013] Furthermore, in step (2), when the robot and the workpiece reach a relatively stationary state, the deviation of the linear velocity and the deviation of the rotational angular velocity of the robot and the workpiece in the reference coordinate system are both controlled within a preset range.
[0014] Further, in step (3), the pose of the robot tool end in the reference coordinate system is calculated by the pose of the workpiece coordinate system in the reference coordinate system and the pose of the robot tool end in the workpiece coordinate system through a preset formula. The preset formula is: pose of the robot tool end in the reference coordinate system = pose of the workpiece coordinate system in the reference coordinate system × pose of the robot tool end in the workpiece coordinate system.
[0015] Furthermore, in step (3), the switching of the robot's reference coordinate system to the workpiece coordinate system is achieved through the coordinate system switching module of the robot control system. The coordinate system switching module pre-stores the conversion relationship between the workpiece coordinate system and the reference coordinate system.
[0016] Furthermore, in step (3), when determining the pose deviation between the robot pose and the key point pose on the workpiece in the workpiece coordinate system, the robot pose and the key point pose are transformed into coordinates based on the homogeneous transformation matrix, and the positional deviation of the two in the coordinate axis direction and the attitude deviation around the coordinate axis are calculated.
[0017] Furthermore, the trajectory planning for pose deviation described in steps (1) and (3) adopts joint space trajectory planning or Cartesian space trajectory planning, and the trajectory planning process avoids robot kinematic singularities.
[0018] Furthermore, it also includes a detection step, which is performed before step (1). Specifically, the detection step is as follows: first, a task queue is established; then, the workpieces are identified and the coordinates of each workpiece are obtained, and the rotation information of the disk is obtained through the encoder. The coordinates of each workpiece and the current rotation information of the disk are stored in the task queue as feature information of the corresponding preset task; finally, the task queue is traversed to determine whether each preset task is repeated with the following task in the task queue. If it is repeated, the preset task is discarded; if it is not repeated, the preset task is inserted into the task queue.
[0019] Furthermore, the rotation information of the disk includes the rotation angle, rotational angular velocity, and rotational angular acceleration of the disk.
[0020] The present invention provides a robot disc conveyor belt following method, which, through a step-by-step design of "pose calibration in reference coordinate system - relative static realization - pose synchronization of key points", combined with coordinate system one, trajectory planning and trajectory superposition technology, realizes a clear and complete disc conveyor belt following model and implementation method, avoids the uncertain following error of existing methods, effectively ensures the following accuracy, and ensures the smooth movement of the robot, adapting to the needs of industrial mass production. Attached Figure Description
[0021] Figure 1 This is a flowchart of a robot disc-shaped conveyor belt following method according to the present invention; Figure 2 This is a schematic diagram showing the positional relationship between the workpiece and the end effector of the robot tool relative to the disc conveyor belt in the robot disc conveyor belt following method of the present invention. Detailed Implementation
[0022] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. The described embodiments are merely some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0023] This solution provides a robot-following method for a disc-shaped conveyor belt. It describes the poses of the robot and the workpiece in the same coordinate system, establishes a pose constraint model for the robot and the workpiece, and ultimately achieves dynamic following of the workpiece on the disc-shaped conveyor belt by the robot. The specific steps include: (a) Testing steps Before executing the robot's following action, a detection step is first performed to obtain relevant workpiece information and establish a task queue, as follows: 1. Establish a task queue: Initialize a task queue for storing follow-up tasks through the robot control system. The task queue supports task insertion, deletion and traversal operations to ensure that follow-up tasks can be managed in an orderly manner. 2. Feature Information Acquisition: Workpieces on the disc-shaped conveyor belt are identified using a vision sensor to accurately acquire the coordinate data of each workpiece. Simultaneously, the rotation information of the disc is acquired in real time through an encoder installed on the disc. The rotation information of the disc includes the rotation angle, rotational angular velocity, and rotational angular acceleration, which can comprehensively reflect the motion state of the disc. The coordinate data of each workpiece and the current disc rotation information are used as feature information for the corresponding preset task, and a unique task identifier is assigned to each preset task to facilitate the differentiation and management of subsequent tasks. 3. Filter the task queue: Iterate through the preset tasks with stored feature information, compare the task identifier of each preset task with the task identifier of the existing follower tasks in the task queue one by one, and determine whether there is a duplicate; if there is a duplicate, discard the preset task to avoid repeated execution and waste of resources; if there is no duplicate, insert the preset task into the task queue to form a list of follower tasks to be executed, ensuring that the follower tasks are executed in order.
[0024] (ii) Execution steps, such as Figure 1 and Figure 2 As shown Step (1): Move the robot to the current pose of the workpiece. Retrieve follow-up tasks from the task queue in order, and perform the following operations in an orderly manner based on the feature information of each follow-up task: Step (101) Refresh workpiece pose: Control the vision sensor to continuously collect image information of the workpiece at a preset fixed frequency, process the collected images through a mature image recognition algorithm, accurately extract the position coordinates and posture angle of the workpiece, and ensure that the obtained workpiece pose information is consistent with the actual state. Step (102) Obtain robot pose: The angle information of each joint is collected in real time by the position sensors installed on each joint of the robot. Combined with the preset robot kinematics forward solution model, the pose data of the robot tool end in the current state is accurately calculated, which provides a basis for subsequent pose comparison. Step (103) Coordinate transformation: The refreshed workpiece pose and the acquired robot pose are transformed into a preset reference coordinate system for unified description. The reference coordinate system takes the center of the disc-shaped conveyor belt as the origin, the X-axis is radially outward along the disc, the Y-axis is tangential to the disc, and the Z-axis is perpendicular to the upper surface of the disc. The difference in pose description is eliminated by the unified coordinate system, which facilitates subsequent deviation calculation. Step (104) Calculate the pose deviation: Based on the workpiece pose and robot pose described uniformly under the reference coordinate system, calculate the position deviation of the two in each coordinate axis direction and the attitude deviation around each coordinate axis to fully reflect the pose difference between the robot and the workpiece. Step (105) Trajectory planning and motion control: Using joint space trajectory planning or Cartesian space trajectory planning, the pose deviation calculated above is scientifically and reasonably planned. During the planning process, the robot kinematic singularities are strictly avoided to prevent the robot from getting stuck during the movement. According to the planned smooth motion trajectory, the robot control system accurately controls the movement of each joint of the robot so that the end of the robot tool can accurately move to the current pose of the workpiece.
[0025] Step (2): Keep the robot and the workpiece relatively stationary. Step (201) Motion parameter acquisition: Based on the disk rotation information collected by the encoder, combined with the workpiece's pose data, the linear velocity and angular velocity of the workpiece in the reference coordinate system are calculated in real time through the preset calculation model to accurately grasp the motion state of the workpiece. Step (202) Acceleration motion planning: Based on the curvature of the workpiece's arc-shaped moving path, the workpiece's current linear velocity and rotational angular velocity, and taking into account the robot's motion performance parameters, scientifically determine the magnitude and acceleration time of the robot's acceleration motion to ensure that the acceleration is within the robot's load-bearing range and avoid excessive acceleration that could cause the robot to vibrate or be damaged; the acceleration trajectory is planned using an S-curve acceleration and deceleration algorithm to further ensure the smoothness of the robot's motion. Step (203) Execute acceleration motion: Control the robot to strictly execute the planned acceleration motion along the arc-shaped movement path of the workpiece. During the acceleration motion, the robot's motion speed data is collected in real time by the sensor and compared and adjusted in real time with the motion speed of the workpiece. When the acceleration motion ends, the robot's linear velocity and rotational angular velocity in the reference coordinate system are consistent with the workpiece, and the robot and the workpiece reach a relatively stationary state, laying the foundation for subsequent pose synchronization.
[0026] Step (3): Synchronize the poses of the key points on the robot and the workpiece. Step (301) Coordinate system switching: Through the coordinate system switching module of the robot control system, the robot's reference coordinate system is accurately switched from the reference coordinate system to the workpiece coordinate system. The workpiece coordinate system takes the geometric center of the workpiece as the origin, and its coordinate axis direction is consistent with the coordinate axis direction of the reference coordinate system, so that the robot and the workpiece maintain a relatively static pose constraint relationship in the workpiece coordinate system. Step (302) Key point pose acquisition: Set key points on the workpiece in advance according to actual operation requirements, such as the gripping point and machining reference point of the workpiece. Accurately acquire the pose data of the key point in the workpiece coordinate system through vision sensor or preset workpiece model data. Step (303) Calculate key point deviation: Based on the robot pose and key point pose in the workpiece coordinate system, accurately calculate the pose deviation between the two, including position deviation and attitude deviation, and clarify the pose difference between the robot and the key points. Step (304) Deviation trajectory planning: Perform targeted trajectory planning for the pose deviations of the above key points to generate a deviation trajectory. The motion speed of the deviation trajectory is adapted according to the relative static state of the robot and the workpiece to ensure that the established relative static relationship between the two is not destroyed. Step (305) Trajectory superposition and synchronization control: Based on the pose constraints of the robot and the workpiece being relatively stationary, the planned deviation trajectory is superimposed on the robot's basic motion trajectory to form the robot's final motion trajectory; the robot control system controls the robot to move precisely according to the final motion trajectory, so that the robot tool end moves accurately to the key point pose on the workpiece. Step (306) Pose Calculation and Verification: Calculate the actual pose of the robot tool end effector in the reference coordinate system using a preset formula: where the preset formula is "the pose of the robot tool end effector in the reference coordinate system". = Workpiece coordinate system pose in the reference coordinate system × Position of the robot tool end effector in the workpiece coordinate system ", as shown in the following formula: , Among them, the pose of the workpiece coordinate system in the reference coordinate system The pose of the robot tool end effector in the workpiece coordinate system is calculated using the disk rotation information obtained from the encoder and the initial value of the workpiece pose. It is obtained by combining the forward kinematics of the robot with coordinate system transformation; The calculated actual pose is compared with the preset key point pose to ensure that the robot and the key points on the workpiece maintain pose synchronization.
[0027] (III) Finishing Steps After the robot completes the following operation on the current workpiece (such as grasping, processing, inspection, etc.), the following task is deleted from the task queue and the task queue resources are released. It is determined whether the task queue is empty. If it is not empty, it returns to step (1) to continue to execute the next following task and realize continuous following operation. If it is empty, it controls the robot to move to the preset standby position and completes the following operation.
[0028] The method in this embodiment enables the robot to accurately and stably follow parts on a disc-shaped conveyor belt. During the following process, the robot moves smoothly and has high synchronization accuracy with the pose of key points of the parts. This effectively avoids the uncertain following errors present in the prior art, fully meets the requirements for following accuracy and stability in industrial production, and has a clear process and strong operability, making it directly applicable to industrial production scenarios.
[0029] The above description is merely illustrative of the embodiments of the present invention and is not intended to limit the present invention. For those skilled in the art, any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A method for robot following a disc-shaped conveyor belt, characterized in that, Includes the following steps: Step (1): Move the robot to the current pose of the workpiece. Refresh the workpiece pose, obtain the robot pose, convert the workpiece pose and robot pose to a reference coordinate system, determine the pose deviation between the robot pose and the workpiece pose in the reference coordinate system, perform trajectory planning on the pose deviation, and control the robot to move to the current pose of the workpiece in the reference coordinate system. Step (2): Keep the robot and the workpiece relatively stationary. Plan an acceleration motion and control the robot to move along the arc-shaped movement path of the workpiece. When the acceleration motion ends, the robot's linear velocity and angular velocity in the reference coordinate system are consistent with the workpiece, and the robot and the workpiece reach a state of relative stillness. Step (3): Synchronize the poses of the key points on the robot and the workpiece. First, switch the robot's reference coordinate system to the workpiece coordinate system to maintain a relatively static pose constraint relationship between the robot and the workpiece. Then, determine the pose deviation between the robot's pose and the pose of key points on the workpiece in the workpiece coordinate system, and perform trajectory planning on the pose deviation. Finally, based on the pose constraint of the robot and the workpiece being relatively static, superimpose the deviation trajectory of the robot's pose and the pose of key points on the workpiece, so that the robot moves to the pose of the key points on the workpiece, and finally maintains a pose synchronization state with the key points on the workpiece.
2. The robot disc-shaped conveyor belt following method according to claim 1, characterized in that, The initial value of the workpiece pose in step (1) is obtained by real-time detection of the workpiece on the disc-shaped conveyor belt by a vision sensor. The vision sensor collects image information of the workpiece at a preset frequency and extracts the pose data of the workpiece based on an image recognition algorithm.
3. The robot disc-shaped conveyor belt following method according to claim 1, characterized in that, When planning an acceleration motion in step (2), the magnitude of the acceleration and the acceleration time are determined based on the curvature of the workpiece's arc-shaped moving path, the workpiece's current linear velocity and rotational angular velocity, so that the robot can smoothly reach a motion state consistent with the workpiece.
4. A robot disc-shaped conveyor belt following method according to claim 1 or 3, characterized in that, In step (2), when the robot and the workpiece reach a relatively stationary state, the deviation of the linear velocity and the deviation of the rotational angular velocity of the robot and the workpiece in the reference coordinate system are both controlled within the preset range.
5. A robot disc-shaped conveyor belt following method according to claim 1, characterized in that, In step (3), the pose of the robot tool end in the reference coordinate system is calculated by the pose of the workpiece coordinate system in the reference coordinate system and the pose of the robot tool end in the workpiece coordinate system through a preset formula. The preset formula is: pose of the robot tool end in the reference coordinate system = pose of the workpiece coordinate system in the reference coordinate system × pose of the robot tool end in the workpiece coordinate system.
6. The robot disc-shaped conveyor belt following method according to claim 1, characterized in that, The step (3) involves switching the robot's reference coordinate system to the workpiece coordinate system through the coordinate system switching module of the robot control system. The coordinate system switching module pre-stores the conversion relationship between the workpiece coordinate system and the reference coordinate system.
7. The robot disc-shaped conveyor belt following method according to claim 1, characterized in that, In step (3), when determining the pose deviation between the robot pose and the key point pose on the workpiece in the workpiece coordinate system, the robot pose and the key point pose are transformed into coordinates based on the homogeneous transformation matrix, and the position deviation of the two in the coordinate axis direction and the attitude deviation around the coordinate axis are calculated.
8. A robot disc-shaped conveyor belt following method according to claim 1, characterized in that, The trajectory planning for pose deviation described in steps (1) and (3) adopts joint space trajectory planning or Cartesian space trajectory planning, and avoids robot kinematic singularities during the trajectory planning process.
9. A robot disc-shaped conveyor belt following method according to claim 1, characterized in that, It also includes a detection step, which is performed before step (1). Specifically, the detection step is as follows: first, a task queue is established; then, the workpiece is identified and the coordinates of each workpiece are obtained. At the same time, the rotation information of the disk is obtained through the encoder. The coordinates of each workpiece and the current rotation information of the disk are stored in the task queue as feature information of the corresponding preset task; finally, the task queue is traversed to determine whether each preset task is repeated with the following task in the task queue. If it is repeated, the preset task is discarded. If it is not repeated, the preset task is inserted into the task queue.
10. A robot disc-shaped conveyor belt following method according to claim 9, characterized in that, The rotation information of the disk includes the rotation angle, rotational angular velocity, and rotational angular acceleration.