Dynamic error compensation method of composite robot based on visual feedback

Abstract

The application discloses a kind of compound robot dynamic error compensation methods based on visual feedback, after compound robot reaches the set station, the initial position of manipulator is acquired using the first camera on the manipulator, and then the static error between the initial position of manipulator and theoretical position is acquired and the static error is compensated, the static error of manipulator can be eliminated;After the initial position of AGV is acquired using the second camera set on AGV, the manipulator moves according to the preset moving track and posture;During the movement of manipulator, the position error between AGV and initial position caused by the rotational inertia of manipulator is acquired by taking a photo of positioning label using the second camera at the set time step Δt, after the position error of AGV exceeds the set threshold, the dynamic error compensation is carried out to the manipulator, until the manipulator reaches the end position;In this way, by compensating the initial static error of manipulator and the dynamic error in the moving process, the working precision can be effectively improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of automatic control, and specifically relates to a dynamic error compensation method of a composite robot based on visual feedback. Background Art

[0002] Composite robots have become one of the important tools for industrial automation, especially in the assembly of parts. Composite robots are generally composed of AGV (Automated Guided Vehicle) and a robotic arm, and the robotic arm is installed on the AGV. During the operation of the composite robot, the AGV is used to move to the specified position, and then the robotic arm is driven to perform operations such as grabbing, assembling, transporting, loading and unloading of materials. However, since the robotic arm generates angular momentum when it rotates, it exerts a reverse force on the AGV. Under the action of this reverse force, the AGV will shake or even move relative to the ground, which will eventually affect the operation accuracy of the composite robot.

[0003] Chinese patent publication number CN109591051B discloses a method for compensating the operating accuracy of a composite robot. This method pre-sets a database table of correspondences between specific actions performed by the composite robot and compensation values. When the composite robot performs a specific action, the corresponding compensation value is called to control the composite robot for error compensation. While this method can improve operating accuracy to a certain extent, it still has the following shortcomings:

[0004] 1) The positioning error between the composite robot and the workstation is not considered: Each time the composite robot moves to the corresponding workstation, there is a positioning error. The positioning error between the composite robot and the same workstation at different times and between the composite robot and different workstations are not the same. However, this method can only perform specific actions according to the set standards and compensate for the error according to the corresponding compensation value. It does not take into account the positioning error between the composite robot and the workstation itself.

[0005] 2) The movement error of the AGV is caused by the angular momentum of the robot arm during rotation, which is related to the mass, center of mass position and angular velocity of the robot arm. During the rotation of the robot arm, the center of mass position and angular velocity of the robot arm will continue to change. Therefore, if the compensation value is to be the same each time, the curve of the change of the center of mass position over time and the curve of the change of the angular velocity over time must be completely consistent each time the robot arm performs the corresponding specific action. However, due to the influence of factors such as the working environment of different workstations and industrial control errors, it is more difficult to maintain complete consistency when performing a specific action each time. Summary of the Invention

[0006] In view of this, the purpose of the present invention is to provide a dynamic error compensation method for a composite robot based on visual feedback, which can not only compensate for the initial static error of the robotic arm, but also obtain the dynamic error of the AGV in real time and realize dynamic compensation of the error to improve the operation accuracy.

[0007] In order to achieve the above object, the present invention provides the following technical solutions:

[0008] A method for dynamic error compensation of a composite robot based on visual feedback comprises the following steps:

[0009] Step 1: Drive the composite robot to the set workstation and stop;

[0010] Step 2: Use the first camera set on the robotic arm to obtain the initial position of the robotic arm, and obtain the static error compensation value Δe of the robotic arm based on the positioning error between the initial position of the robotic arm and the theoretical position. s ;

[0011] Using a second camera provided on the AGV to obtain image information of a positioning label corresponding to the workstation, and using the image information of the positioning label to obtain an initial position of the AGV;

[0012] Step 3: Drive the robotic arm to move and perform static error compensation and dynamic error compensation on the robotic arm. The method for performing dynamic error compensation on the robotic arm is as follows:

[0013] 31): At a set time interval Δt from the last photo taken by the second camera, the second camera is used to obtain image information of the positioning label corresponding to the workstation;

[0014] 32): Obtain the position error Δe between the actual position and the initial position of the AGV at the current moment based on the image information of the positioning tag d ;

[0015] 33): Determine whether the position error is greater than or equal to the set threshold: If so, the position error is used to obtain the current dynamic error compensation value for the robot arm, and the dynamic error compensation is performed on the robot arm; if not, no dynamic error compensation is performed;

[0016] 34): Determine whether the robotic arm has reached the end position: if so, complete dynamic error compensation; if not, drive the robotic arm to continue moving according to the set trajectory and execute step 31).

[0017] Furthermore, in step 32), the method for obtaining the positioning error of the AGV at the current moment based on the image information of the positioning tag is:

[0018] 321) Processing image information to obtain pixel coordinates of feature points;

[0019] 322) Construct a moving coordinate system that moves synchronously with the AGV, convert the pixel coordinate system to the moving coordinate system, and obtain the actual coordinate value of the feature point in the moving coordinate system;

[0020] 323) Using the actual coordinate values ​​and theoretical coordinate values ​​of the feature points in the mobile coordinate system, the position error between the actual position and the initial position of the AGV at the current moment is calculated.

[0021] Furthermore, in step 321), the image information is processed as follows:

[0022] 3211) grayscale processing of image information using enhanced average value;

[0023] 3212) performing filtering on the grayscale processed image information and retaining feature point information in the image information;

[0024] 3213) performing binarization processing on the image;

[0025] 3214) Create a pixel coordinate system and obtain the pixel coordinates of the feature point in the pixel coordinate system.

[0026] Furthermore, in step 33), the method for obtaining the dynamic error compensation value of the robot arm using the position error is:

[0027] ΔU=KΔE

[0028] Where ΔU and ΔE represent the frequency domain transformations of Δu and Δe, respectively; Δu represents the time domain output of the dynamic error compensation controller; Δe represents the comprehensive position error between the actual position and the initial position of the AGV, and satisfies Δe=Δe s +Δe d ; K represents the dynamic error compensation controller.

[0029] Furthermore, the dynamic error compensation controller K adopts a PID controller, a fuzzy PID controller, a robust controller, an adaptive controller or a neural network controller.

[0030] The beneficial effects of the present invention are:

[0031] The present invention discloses a dynamic error compensation method for a composite robot based on visual feedback. First, after the composite robot arrives at a set workstation, the first camera on the robotic arm is used to obtain the initial position of the robotic arm, and then the static error between the initial position and the theoretical position of the robotic arm is obtained and the static error is compensated. This method can eliminate different static errors between the composite robot arriving at the same workstation at different times and between the composite robot and different workstations. After the second camera on the AGV is used to obtain the initial position of the AGV, the robotic arm is moved according to a preset movement trajectory and posture. During the movement of the robotic arm, the second camera is used to take a picture of the positioning tag at a set time step Δt to obtain the position error between the AGV and the initial position caused by the rotational inertia of the robotic arm. After the position error of the AGV exceeds a set threshold, the robotic arm is dynamically compensated until the robotic arm reaches the end position. In this way, by compensating for the initial static error of the robotic arm and the dynamic error during the movement process, the operation accuracy can be effectively improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to make the purpose, technical solutions and beneficial effects of the present invention more clear, the present invention provides the following drawings for illustration:

[0033] Figure 1 This is a flow chart of an embodiment of the method for dynamic error compensation of a composite robot based on visual feedback according to the present invention. DETAILED DESCRIPTION

[0034] The present invention will be further described below with reference to the accompanying drawings and specific embodiments so that those skilled in the art can better understand the present invention and implement it. However, the embodiments are not intended to limit the present invention.

[0035] like Figure 1 As shown, the dynamic error compensation method of the composite robot based on visual feedback of this embodiment includes the following steps.

[0036] Step 1: Drive the composite robot to the set workstation and stop.

[0037] Step 2: Use the first camera set on the robotic arm to obtain the initial position of the robotic arm, and obtain the static error compensation value Δe of the robotic arm based on the positioning error between the initial position of the robotic arm and the theoretical position. s .

[0038] The second camera provided on the AGV is used to obtain image information of the positioning label corresponding to the workstation, and the image information of the positioning label is used to obtain the initial position of the AGV.

[0039] Step 3: Drive the robotic arm to move and perform static error compensation and dynamic error compensation on the robotic arm.

[0040] Among them, the method for dynamic error compensation of the robotic arm is:

[0041] 31) At a set time interval Δt from the last photo taken by the second camera, the second camera is used to obtain image information of the positioning label corresponding to the workstation.

[0042] 32): Obtain the position error between the actual position and the initial position of the AGV at the current moment based on the image information of the positioning tag. Specifically, in this embodiment, the method for obtaining the positioning error of the AGV at the current moment based on the image information of the positioning tag is:

[0043] 321) Process the image information to obtain the pixel coordinates of the feature points. Specifically, the image information processing method of this embodiment is as follows:

[0044] 3211) grayscale processing of image information using enhanced average value;

[0045] 3212) performing filtering on the grayscale processed image information and retaining feature point information in the image information;

[0046] 3213) performing binarization processing on the image;

[0047] 3214) Create a pixel coordinate system and obtain the pixel coordinates of the feature point in the pixel coordinate system.

[0048] 322) Construct a moving coordinate system that moves synchronously with the AGV, convert the pixel coordinate system to the moving coordinate system, and obtain the actual coordinate value of the feature point in the moving coordinate system.

[0049] 323) Using the actual coordinate value and theoretical coordinate value of the feature point in the mobile coordinate system, the position error Δe between the actual position and the initial position of the AGV at the current moment is calculated. d .

[0050] 33): Determine whether the position error is greater than or equal to the set threshold: If so, the position error is used to obtain the current dynamic error compensation value of the robot arm and the dynamic error compensation is performed on the robot arm; if not, no dynamic error compensation is performed. The method for obtaining the dynamic error compensation value of the robot arm using the position error is:

[0051] ΔU=KΔE

[0052] Where ΔU and ΔE represent the frequency domain transformations of Δu and Δe, respectively; Δu represents the time domain output of the dynamic error compensation controller; Δe represents the comprehensive position error between the actual position and the initial position of the AGV, and satisfies Δe=Δe s +Δe dK represents a dynamic error compensation controller. The dynamic error compensation controller can be a PID controller, a fuzzy PID controller, a robust controller, an adaptive controller, or a neural network controller. In this embodiment, the error compensation controller is a PID controller.

[0053] Specifically, the selection of the threshold can be set according to the accuracy requirements of different scenarios, which will not be described in detail.

[0054] 34): Determine whether the robotic arm has reached the end position: if so, complete dynamic error compensation; if not, drive the robotic arm to continue moving according to the set trajectory and execute step 31).

[0055] The above embodiments are merely preferred embodiments for the purpose of fully illustrating the present invention, and the scope of protection of the present invention is not limited thereto. Equivalent substitutions or modifications made by those skilled in the art based on the present invention are within the scope of protection of the present invention. The scope of protection of the present invention shall be subject to the claims.

Claims

1. A dynamic error compensation method for a composite robot based on visual feedback, characterized by: The steps include: Step 1: Drive the composite robot to the set workstation and stop; Step 2: Use the first camera set on the robotic arm to obtain the initial position of the robotic arm, and obtain the static error compensation value Δe of the robotic arm based on the positioning error between the initial position of the robotic arm and the theoretical position. s ; Using a second camera provided on the AGV to obtain image information of a positioning label corresponding to the workstation, and using the image information of the positioning label to obtain an initial position of the AGV; Step 3: Drive the robotic arm to move and perform static error compensation and dynamic error compensation on the robotic arm. The method for performing dynamic error compensation on the robotic arm is as follows: 31) At a set time interval Δt from the last time the second camera took a picture, the second camera is used to obtain image information of the positioning label corresponding to the workstation; 32): Obtain the position error Δe between the actual position and the initial position of the AGV at the current moment based on the image information of the positioning tag d ; 33): Determine whether the position error is greater than or equal to the set threshold: If so, the position error is used to obtain the current dynamic error compensation value for the robot arm, and the dynamic error compensation is performed on the robot arm; if not, no dynamic error compensation is performed; 34): Determine whether the robotic arm has reached the end position: if so, complete dynamic error compensation; if not, drive the robotic arm to continue moving according to the set trajectory and execute step 31).

2. The method for dynamic error compensation of a composite robot based on visual feedback according to claim 1, characterized in that: In step 32), the method for obtaining the positioning error of the AGV at the current moment based on the image information of the positioning tag is: 321) Processing image information to obtain pixel coordinates of feature points; 322) Construct a moving coordinate system that moves synchronously with the AGV, convert the pixel coordinate system to the moving coordinate system, and obtain the actual coordinate value of the feature point in the moving coordinate system; 323) Using the actual coordinate value and theoretical coordinate value of the feature point in the mobile coordinate system, the position error Δe between the actual position and the initial position of the AGV at the current moment is calculated. d .

3. The method for dynamic error compensation of a composite robot based on visual feedback according to claim 2, characterized in that: In step 321), the image information is processed as follows: 3211) grayscale processing of image information using enhanced average value; 3212) performing filtering on the grayscale processed image information and retaining feature point information in the image information; 3213) performing binarization processing on the image; 3214) Create a pixel coordinate system and obtain the pixel coordinates of the feature point in the pixel coordinate system.

4. The method for dynamic error compensation of a composite robot based on visual feedback according to claim 1, characterized in that: In step 33), the method for obtaining the dynamic error compensation value of the robot arm using the position error is: ΔU=KΔE Where ΔU and ΔE represent the frequency domain transformations of Δu and Δe, respectively; Δu represents the time domain output of the dynamic error compensation controller; Δe represents the comprehensive position error between the actual position and the initial position of the AGV, and satisfies Δe=Δe s +Δe d ; K represents the error compensation controller.

5. The method for dynamic error compensation of a composite robot based on visual feedback according to claim 4, characterized in that: The error compensation controller adopts a PID controller, a fuzzy PID controller, a robust controller, an adaptive controller or a neural network controller.

Images