WO2024113216A1 - 一种工业模具智能制造机器人高精度抓取方法 - Google Patents
一种工业模具智能制造机器人高精度抓取方法 Download PDFInfo
- Publication number
- WO2024113216A1 WO2024113216A1 PCT/CN2022/135360 CN2022135360W WO2024113216A1 WO 2024113216 A1 WO2024113216 A1 WO 2024113216A1 CN 2022135360 W CN2022135360 W CN 2022135360W WO 2024113216 A1 WO2024113216 A1 WO 2024113216A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- grasping
- robot
- intelligent manufacturing
- camera
- precision
- Prior art date
Links
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 54
- 238000000034 method Methods 0.000 title claims abstract description 53
- 230000003044 adaptive effect Effects 0.000 claims abstract description 13
- 230000007704 transition Effects 0.000 claims description 13
- 238000003384 imaging method Methods 0.000 claims description 10
- 238000012545 processing Methods 0.000 claims description 6
- 238000005516 engineering process Methods 0.000 claims description 4
- 238000012546 transfer Methods 0.000 claims description 2
- 238000011160 research Methods 0.000 abstract description 3
- 230000033001 locomotion Effects 0.000 description 5
- 230000009466 transformation Effects 0.000 description 4
- 239000011159 matrix material Substances 0.000 description 3
- 238000013016 damping Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 101001121408 Homo sapiens L-amino-acid oxidase Proteins 0.000 description 1
- 102100026388 L-amino-acid oxidase Human genes 0.000 description 1
- 230000001133 acceleration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 230000009193 crawling Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000005484 gravity Effects 0.000 description 1
- 238000005312 nonlinear dynamic Methods 0.000 description 1
- 238000004806 packaging method and process Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
- 238000003466 welding Methods 0.000 description 1
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J9/00—Programme-controlled manipulators
Definitions
- the invention belongs to the technical field of robot grasping, and in particular relates to a high-precision grasping method of an industrial mold intelligent manufacturing robot.
- Robots have been widely used in welding, palletizing, packaging and other industrial industries, which can not only improve labor productivity but also reduce labor costs.
- Industrial mold intelligent manufacturing robots have the advantages of precision, high flexibility, high speed, high rigidity, and light weight, and are suitable for grasping work. Labor costs continue to increase with economic development. In order to reduce labor costs, more and more companies are beginning to use robots. In the context of continuous progress and development of science and technology, industrial mold intelligent manufacturing robots have become an indispensable part of the manufacturing industry. Therefore, it is of great significance to study the high-precision grasping method of robots.
- Zou Yu et al. proposed a robot grasping control method based on the combination of 3D point cloud depth information and centroid distance.
- the data was collected through the Kinect sensor, denoised and segmented to obtain the boundary point cloud of the grasped target, the angle between the grasping point and the normal vector of the boundary point cloud in the straight line was obtained, and a comprehensive evaluation function was constructed to obtain the optimal grasping position and complete the robot grasping control.
- this method controls the robot to grasp, it cannot accurately track the changes in the robot joints, the tracking error is large, and there is a problem of low grasping accuracy.
- Ge Junyan [4] et al. proposed a robot grasping method based on a 3D detection network.
- the acquired image is input into a convolutional neural network to identify the object information, and the 3D bounding box of the target is constructed based on the acquired information.
- the center point of the target is obtained, and the grasping strategy is designed to realize the robot grasping control.
- this method plans the robot's grasping path, the shortest grasping path and the average grasping path are both long, which increases the time taken by the robot to grasp and has the problem of low grasping efficiency.
- the purpose of the present invention is to provide a high-precision grasping method for an industrial mold intelligent manufacturing robot to solve the problems existing in the above-mentioned prior art.
- the present invention provides a high-precision grasping method for an industrial mold intelligent manufacturing robot, comprising:
- the process of calibrating the camera includes: converting the world coordinate system to the camera coordinate system, obtaining the imaging plane coordinate system based on the camera coordinate system, processing the imaging plane coordinate system based on the radial distortion value to obtain corrected coordinates, and obtaining image coordinates corresponding to the projection center based on the corrected coordinates, the image sensor pixel size and the vertical projection of the projection center in the imaging plane coordinate system.
- the process of obtaining robot information and target point information includes: obtaining images of the robot and the target workpiece, image acquisition times, and positions of the robot and the target workpiece based on an industrial smart camera, a preset acquisition interval, and a speed encoder, wherein the industrial smart camera and the speed encoder are installed on a conveyor belt, and the industrial smart camera is a camera for calibration processing.
- the acquisition interval is obtained based on a maximum length of a camera field of view and a target workpiece.
- the process of acquiring the target workpiece information further includes: constructing sets of two adjacent images respectively, and determining whether there is duplication by comparing elements in the sets pairwise, wherein the elements in the sets are coordinate information of the images.
- the process of obtaining the grasping path includes: determining the coordinates of the target workpiece grasping point and the target workpiece control point, obtaining the transition point on the conveyor belt when it is lifted to a preset height, obtaining the remaining passing points based on the transition radius, and obtaining the grasping path based on the principles of path smoothness and arc transition corners, combining the obtained points, wherein the transition radius is obtained based on the algorithm of inserting the transition arc radius into the transfer point.
- the process of obtaining the grasping position includes: obtaining the grasping position based on the position corresponding to the target workpiece and the conveyor belt speed and using a binary search method.
- an adaptive coordination controller model is designed based on the grasping trajectory, the grasping position and the master-slave multi-technology.
- the present invention proposes a high-precision grasping method for an industrial mold intelligent manufacturing robot.
- the grasping trajectory and grasping points are determined according to the acquired robot position information, and an adaptive coordination controller is designed to achieve high-precision grasping of the industrial mold intelligent manufacturing robot, providing a reliable research basis for multi-target grasping of the industrial mold intelligent manufacturing robot and other research in the field.
- FIG1 is a schematic diagram of a grabbing path in an embodiment of the present invention.
- FIG. 2 is a schematic diagram of a method flow in an embodiment of the present invention.
- this embodiment provides a high-precision grasping method of an industrial mold intelligent manufacturing robot, including:
- the camera In order to obtain the position of the industrial mold intelligent manufacturing robot and the target point, the camera needs to be calibrated and processed as follows:
- a c (x c , y c , z c ) T and A w (x w , y w , z w ) T are used to represent the points in the camera coordinate system and the world coordinate system respectively.
- the above transformation process is described by the following formula:
- T ( ⁇ , ⁇ , ⁇ ) represents the rotation matrix, where ⁇ , ⁇ , and ⁇ represent the rotation angles around the x-axis, y-axis, and z-axis in the camera coordinate system, respectively;
- U (u x , u y , u z ) T represents the translation vector.
- the calculated imaging plane coordinates usually have errors.
- the main reason for this phenomenon is the existence of lens distortion. Therefore, the high-precision grasping method of the industrial mold intelligent manufacturing robot corrects the lens distortion through the following formula:
- (c, d) represents the coordinates corresponding to the projection center; D y and D x both represent the pixel size of the image sensor; V y and V x both represent the vertical projection of the projection center in the imaging plane.
- the camera calibration is completed to solve the lens distortion problem.
- the calibrated camera is used to obtain relevant information between the industrial mold intelligent manufacturing robot and the target point. This section comprehensively plans the grasping strategy.
- the high-precision grasping method of the industrial mold intelligent manufacturing robot sets the speed encoder in the conveyor belt, and obtains the position coordinates ( xn , yn , Mn ) of the target workpiece at any time according to the number of feedback pulses obtained.
- the position coordinates (x, y) of the target during the operation of the conveyor belt can be obtained at any time:
- Lc represents the distance traveled per unit pulse number
- Mn represents the pulse value of the speed encoder at the time of image acquisition
- M represents the pulse value of the speed encoder at the current moment.
- the conveyor belt usually works continuously, so the target workpiece will constantly enter the camera's field of view.
- the camera In order to obtain the position of the target workpiece, the camera needs to continuously collect images of the target workpiece.
- the collection interval needs to be determined. If the interval is too short, the number of image processing will increase. If the interval is too long, it is easy to miss the target.
- the camera's field of view is represented by a ⁇ b.
- the high-precision grasping method of the industrial mold intelligent manufacturing robot uses the speed encoder feedback to complete the control, and the acquisition interval of the industrial intelligent camera is set to the transmission distance ds per movement.
- the conveyor belt and the target workpiece are in a stationary state, so the repeatedly photographed target workpiece position information coordinate y remains consistent after image processing, and the x coordinate changes, with a difference of ds.
- the maximum length of the target workpiece in the x direction is l, and at this time:
- the phenomenon of repeated workpiece position information may appear in two adjacent images.
- the above sets are composed of the coordinate information in the first and second images respectively. By comparing the elements one by one, it is determined whether there are duplicate coordinates in set S and set B:
- ⁇ represents the pairwise comparison matrix; when the value of the parameter ⁇ i,j is zero, it indicates that repeated acquisition has occurred and the two coordinate points are repeated.
- the main goal of high-precision grasping of industrial mold intelligent manufacturing robots is to accurately control the robot to reach the placement point and grasping point, reduce the robot's movement time, and ensure the smoothness of the robot's movement trajectory.
- T represents the time it takes the robot to move from (x 0 ,y 0 ) to (x d ,y d ).
- the high-precision grasping method of the industrial mold intelligent manufacturing robot combines the conveyor belt speed on the basis of solving the equation by dichotomy to obtain the position corresponding to the grasping point:
- xn0 represents the position of the target workpiece when the robot starts moving.
- the high-precision grasping method of the industrial mold intelligent manufacturing robot is designed based on a one-master-multiple-slave adaptive collaborative controller model according to the planned grasping trajectory and grasping point position obtained in the above process to achieve high-precision grasping work of the industrial mold intelligent manufacturing robot.
- G N represents the force generated by the manufacturing environment
- a N represents the velocity of the joint of the industrial mold intelligent manufacturing robot
- Q N (a N ) represents the positive definite inertia matrix
- ⁇ N represents the robot torque
- h N (a N ) represents the gravity term
- G h represents the force applied to the robot by the operator.
- the high-precision grasping method of the industrial mold intelligent manufacturing robot establishes the adaptive collaborative controller model of the tool mold intelligent manufacturing robot, and the expression is as follows:
- ⁇ m represents the damping coefficient corresponding to the main manufacturing robot
- k represents the proportional index
- fim (t) represents the delay time of the manufacturing robot
- ⁇ z represents the following structure
- ⁇ z [N] represents the number of adjacent industrial mold intelligent manufacturing robots
- tfim (t) represents the delay time between adjacent robots, all of which are measured values
- ⁇ m represents the torque of the main manufacturing robot and am represents the velocity of the main manufacturing robot.
- ⁇ s represents the damping coefficient corresponding to the manufacturing robot.
- ⁇ g represents the leadership structure from which the robot is manufactured.
- the high-precision grasping method of the industrial mold intelligent manufacturing robot uses the adaptive collaborative controller constructed by the above process to complete the high-precision grasping control of the industrial mold intelligent manufacturing robot.
Landscapes
- Engineering & Computer Science (AREA)
- Robotics (AREA)
- Mechanical Engineering (AREA)
- Moulds For Moulding Plastics Or The Like (AREA)
Abstract
一种工业模具智能制造机器人高精度抓取方法，包括：标定相机，通过校正相机镜头获得工业模具智能制造机器人和目标点信息，规划工业模具智能制造机器人的抓取轨迹，并确定抓取点位置。基于此，构建自适应协调控制器，结合抓取轨迹和目标抓取点，构建自适应协调控制模型，实现工业模具智能制造机器人的高精度抓取。根据获取的机器人位置信息确定抓取轨迹和抓取点，设计自适应协调控制器，实现工业模具智能制造机器人的高精度抓取，为工业模具智能制造机器人的多目标抓取及领域内其他的研究提供可靠的探究基础。
Description
本发明属于机器人抓取技术领域，特别是涉及一种工业模具智能制造机器人高精度抓取方法。
在焊接、码垛、包装等工业行业中机器人已经得到了广泛的应用，不仅可以提高劳动生产率，还能降低劳动成本。工业模具智能制造机器人具有精准、高柔性、高速、高刚性、质量轻等优点，适用于抓取工作。劳动成本随着经济的发展不断提高，为了降低劳动成本，越来越多的企业开始使用机器人，在科技水平不断进步和发展的背景下，工业模具智能制造机器人成为制造业中不可或缺的一部分，因此研究机器人高精度抓取方法具有重要意义。
邹遇等人提出基于3维点云深度信息和质心距相结合的机器人抓取控制方法。通过Kinect传感器采集数据，对数据实行去噪和分割处理，得到抓取目标的边界点云，获取直线中抓取点与边界点云法向量之间存在的夹角，构建综合评价函数，获取最优抓取位置，完成机器人的抓取控制，该方法控制机器人实行抓取工作时，无法准确跟踪机器人关节的变化情况，跟踪误差较大，存在抓取精度低的问题。葛俊彦
[4]等人提出基于三维检测网络的机器人抓取方法。将获取的图像输入卷积神经网络中，对物体信息识别，根据获取的信息构建目标的三维包围盒，获取目标的中心点，以此设计抓取策略，实现机器人的抓取控制，该方法规划机器人的抓取路径时，获得的最短抓取路径和平均抓取路径均较长，增加了机器人抓取所用的时 间，存在抓取效率低的问题。
发明内容
本发明的目的是提供一种工业模具智能制造机器人高精度抓取方法，以解决上述现有技术存在的问题。
为实现上述目的，本发明提供了一种工业模具智能制造机器人高精度抓取方法，包括：
对相机进行标定处理，通过标定后的相机获得机器人信息与目标工件信息；
基于所述机器人信息与所述目标点信息获得抓取路径与抓取位置；
构建自适应协同控制器模型；
基于所述抓取路径、所述抓取位置及所述自适应协同控制器模型，实现机器人高精度抓取。
可选的，对相机进行标定处理的过程包括：进行对世界坐标系到相机坐标系的转换，基于所述相机坐标系获得成像平面坐标系，基于径向扭曲值对所述成像平面坐标系进行处理，获得校正后的坐标，基于校正后的坐标、图像传感器像元大小与成像平面坐标系内投影中心的垂直投影获得投影中心对应的图像坐标。
可选的，获取机器人信息与目标点信息的过程包括：基于工业智能相机、预设定采集间隔及测速编码器获得机器人与目标工件的图像、图像采集时刻及机器人与目标工件位置，其中，所述工业智能相机与测速编码器安装在传送带上，所述工业智能相机为进行标定处理的相机。
可选的，基于相机视野与目标工件的最大长度获得所述采集间隔。
可选的，目标工件信息的获取过程还包括：分别构建相邻两幅图像的集合，通过集合中的元素两两对比的方式判断是否存在重复，其中，集合中的元素为图像的坐标信息。
可选的，获得抓取路径的过程包括：确定目标工件抓取点与目标工件控制点的坐标，获取在抬起预设高度时传送带上的过渡点，基于过渡半径获得其余经过点，基于路径平滑的原则与弧线过度拐角处的原则，结合获得的点获得抓取路径，其中，过渡半径依据转接点***过渡圆弧半径算法获得。
可选的，获取抓取位置的过程包括：基于目标工件对应的位置、传送带速度，并采用二分法获得抓取位置。
可选的，基于所述抓取轨迹、所述抓取位置及主从多技术设计自适应协调控制器模型。
本发明的技术效果为：
本发明提出一种工业模具智能制造机器人高精度抓取方法，根据获取的机器人位置信息确定抓取轨迹和抓取点，设计自适应协调控制器，实现工业模具智能制造机器人的高精度抓取，为工业模具智能制造机器人的多目标抓取及领域内其他的研究提供可靠的探究基础。
构成本申请的一部分的附图用来提供对本申请的进一步理解，本申请的示意性实施例及其说明用于解释本申请，并不构成对本申请的不当限定。在附图中：
图1为本发明实施例中的抓取路径示意图；
图2为本发明实施例中的方法流程示意图。
需要说明的是，在不冲突的情况下，本申请中的实施例及实施例中的特征可以相互组合。下面将参考附图并结合实施例来详细说明本申请。
需要说明的是，在附图的流程图示出的步骤可以在诸如一组计算机可执行指令的计算机***中执行，并且，虽然在流程图中示出了逻辑顺序，但是在某些情况下，可以以不同于此处的顺序执行所示出或描述的步骤。
实施例一
如图1-2所示，本实施例中提供一种工业模具智能制造机器人高精度抓取方法，包括：
S1.相机标定
在相机成像过程中镜头容易产生切向畸变和径向畸变，为了获取工业模具智能制造机器人和目标点的位置，需要对相机实行标定处理体过程如下：
(1)世界坐标系与相机坐标系之间的转变
可用刚性变换描述两者之间的转变过程，分别用A
c(x
c,y
c,z
c)
T、A
w(x
w,y
w,z
w)
T表示相机坐标系和世界坐标系中存在的点，通过下述公式描述上述转变过程：
A
c＝TA
w+U (1)
式中，T＝(α,β,γ)表示旋转矩阵，其中α、β、γ分别表示在摄像机坐标系中围绕x轴、y轴和z轴的旋转角度；U＝(u
x,u
y,u
z)
T表示平移向量。
计算获取的成像平面坐标通常存在误差，造成这种现象的主要原因是 存在镜头畸变，因此，工业模具智能制造机器人高精度抓取方法通过下式校正镜头畸变：
式中，
代表校正后的坐标；l代表径向扭曲的大小。
(2)成像平面坐标与图像坐标系之间的转换可通过下式得以描述：
式中，(c,d)代表投影中心对应的坐标；D
y、D
x均表示图像传感器像元大小；V
y、V
x均表示在成像平面内投影中心的垂直投影。
S2.抓取策略规划
基于上述过程完成相机标定，解决镜头畸变问题，利用标定后的相机获取工业模具智能制造机器人与目标点的相关信息，本小节全面规划抓取策略。
(一)单目标表示
在传送带上安装工业智能相机，设定采集间隔，获得采集时刻以及目标位置。工业模具智能制造机器人高精度抓取方法将测速编码器设置在传送带中，根据获取的反馈脉冲数获得目标工件在任意时刻中的位置坐标(x
n,y
n,M
n)，通过上述过程可以在任意时刻获取目标在传送带运行过程中的位置坐标(x,y)：
(x,y)＝[x
n+L
c(M-M
n),y
n] (4)
式中，L
c代表单位脉冲数的前进距离；M
n代表测速编码器在图像采集 时刻的脉冲值；M代表测速编码器在当前时刻的脉冲值。
(二)多目标表示
传送带通常情况下是持续工作的，因此目标工件会不断地进入相机视野内，为了获取目标工件的位置，相机需要一直采集目标工件的图像。在采集过程中需要确定采集间隔，间隔太短会增大图像处理数量，间隔过长，容易出现目标漏采的现象。
用a×b表示相机的视野，针对相机采集触发信号，工业模具智能制造机器人高精度抓取方法利用测速编码器反馈完成控制，将工业智能相机的采集间隔设置为传送每移动距离ds。传送带与目标工件之间处于静止状态，因此重复拍摄的目标工件位置信息坐标y经过图像处理后保持一致，x坐标发生变化，相差ds，在x方向中目标工件的最大长度为l，此时存在：
(1)当ds的值大于等于a时，出现漏拍的概率较大；
(2)当ds的值大于等于a+l/2时，获取的目标工件图像完整度较低；
(3)当a-l/2≥ds≥a-l时，重复拍摄概率增大。
通过上述分析工业模具智能制造机器人高精度抓取方法设定ds＝a-l。工件位置信息重复的现象可能会出现在相邻的两幅图像中，为了消除重复的位置坐标，需要逐个对比采集的位置坐标，设定集合S和集合B：
上述集合分别由第一幅和第二幅图像中存在的坐标信息构成，通过元素两两对比，判断集合S和集合B中是否存在重复坐标：
式中，Δ代表两两对比矩阵；当参数ε
i,j的值为零时，表明出现了重复 采集现象，两个坐标点重复。
(三)规划抓取路径
工业模具智能制造机器人执行任务时，可能会受到障碍物的阻碍，“门型”路径是常用的运行路径，工业模具智能制造机器人高精度抓取方法采用弧线过渡拐角处，以提高工业模具机器人在运动过程中的平滑性和运行速度，工业模具智能制造机器人高精度抓取方法生成的笛卡尔空间末端运动路径如图1所示。
首先确定目标抓取点和目标放置点A
1、A
8分别对应的坐标，在抬起高度j的基础上确定传送带中存在的过渡点A
3、A
6，根据其过渡半径r
1、r
2确定该段传送带中存在的经过点A
2、A
4、A
5、A
7，通过上述过程获得工业模具智能制造机器人的抓取放置路径：
工业模具智能制造机器人高精度抓取的主要目标是准确的控制机器人到达放置点和抓取点的同时，减少机器人的运动时间，并且保证机器人运动轨迹的平滑性。
(四)抓取位置
设(x
d,y
d)代表抓取点对应的位置；(x
n,y
n,N
n)代表的目标工件对应的位置，开始工作时，工业模具智能制造机器人末端对应的位置为(x
0,y
0)，编码器在传感器中的当前读数为N
0，从初始点开始工业模具智能制造机器人运动到抓取位置花费的时间t，可通过下式计算得到：
t＝t
s+T (8)
式中，t
s代表补偿时间，T代表机器人从(x
0,y
0)运动到(x
d,y
d)花费的时 间。
工业模具智能制造机器人高精度抓取方法在二分法解方程的基础上结合传送带速度，获得抓取点对应的位置：
式中，x
n0代表目标工件当机器人开始运动时所处的位置。
S3.抓取控制
工业模具智能制造机器人高精度抓取方法根据上述过程获取的规划抓取轨迹和抓取点位置设计基于一主多从的自适应协同控制器模型，实现工业模具智能制造机器人的高精度抓取工作。
通过非线性动力学模型描述一主多从技术：
式中，G
N表示制造环境产生的力；a
N代表工业模具智能制造机器人关节的速率；Q
N(a
N)代表正定惯性矩阵；υ
N代表机器人力矩；
表示向心力；
代表机器人的加速度向量；h
N(a
N)代表重力项；G
h代表操作者施加到机器人上的力。
工业模具智能制造机器人高精度抓取方法建立工具模具智能制造机器人自适应协同控制器模型，表达式如下：
式中，β
m代表主制造机器人对应的阻尼系数；k表示比例指数；f
im(t)代表制造机器人的延迟时间；Δz表示追随结构；Δz
[N]表示相邻工业模具智能制造机器人的数量；t-f
im(t)表示相邻机器人之间存在的延迟时间，均为测 量值，υ
m代表主制造机器人力矩和a
m代表主制造机器人速率。
当i∈Δz时，存在下式：
式中，
表示相邻工业模具制造机器人的数量；β
s代表从制造机器人对应的阻尼系数。
当i∈Δg时，存在下式：
其中，Δg代表从制造机器人的领导结构。
工业模具智能制造机器人高精度抓取方法利用上述过程构建的自适应协同控制器完成工业模具智能制造机器人的高精度抓取控制。
以上所述，仅为本申请较佳的具体实施方式，但本申请的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本申请揭露的技术范围内，可轻易想到的变化或替换，都应涵盖在本申请的保护范围之内。因此，本申请的保护范围应该以权利要求的保护范围为准。
Claims (8)
- 一种工业模具智能制造机器人高精度抓取方法，其特征在于，包括以下步骤：对相机进行标定处理，通过标定后的相机获得机器人信息与目标工件信息；基于所述机器人信息与所述目标点信息获得抓取路径与抓取位置；构建自适应协同控制器模型；基于所述抓取路径、所述抓取位置及所述自适应协同控制器模型，实现机器人高精度抓取。
- 根据权利要求1所述的工业模具智能制造机器人高精度抓取方法，其特征在于，对相机进行标定处理的过程包括：进行对世界坐标系到相机坐标系的转换，基于所述相机坐标系获得成像平面坐标系，基于径向扭曲值对所述成像平面坐标系进行处理，获得校正后的坐标，基于校正后的坐标、图像传感器像元大小与成像平面坐标系内投影中心的垂直投影获得投影中心对应的图像坐标。
- 根据权利要求1所述的工业模具智能制造机器人高精度抓取方法，其特征在于，获取机器人信息与目标点信息的过程包括：基于工业智能相机、预设定采集间隔及测速编码器获得机器人与目标工件的图像、图像采集时刻及机器人与目标工件位置，其中，所述工业智能相机与测速编码器安装在传送带上，所述工业智能相机为进行标定处理的相机。
- 根据权利要求3所述的工业模具智能制造机器人高精度抓取方法，其特征在于，基于相机视野与目标工件的最大长度获得所述采集间隔。
- 根据权利要求1所述的工业模具智能制造机器人高精度抓取方法， 其特征在于，目标工件信息的获取过程还包括：分别构建相邻两幅图像的集合，通过集合中的元素两两对比的方式判断是否存在重复，其中，集合中的元素为图像的坐标信息。
- 根据权利要求1所述的工业模具智能制造机器人高精度抓取方法，其特征在于，获得抓取路径的过程包括：确定目标工件抓取点与目标工件控制点的坐标，获取在抬起预设高度时传送带上的过渡点，基于过渡半径获得其余经过点，基于路径平滑的原则与弧线过度拐角处的原则，结合获得的点获得抓取路径，其中，过渡半径依据转接点***过渡圆弧半径算法获得。
- 根据权利要求1所述的工业模具智能制造机器人高精度抓取方法，其特征在于，获取抓取位置的过程包括：基于目标工件对应的位置、传送带速度，并采用二分法获得抓取位置。
- 根据权利要求1所述的工业模具智能制造机器人高精度抓取方法，其特征在于，基于所述抓取轨迹、所述抓取位置及主从多技术设计自适应协调控制器模型。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/CN2022/135360 WO2024113216A1 (zh) | 2022-11-30 | 2022-11-30 | 一种工业模具智能制造机器人高精度抓取方法 |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/CN2022/135360 WO2024113216A1 (zh) | 2022-11-30 | 2022-11-30 | 一种工业模具智能制造机器人高精度抓取方法 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2024113216A1 true WO2024113216A1 (zh) | 2024-06-06 |
Family
ID=91322608
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2022/135360 WO2024113216A1 (zh) | 2022-11-30 | 2022-11-30 | 一种工业模具智能制造机器人高精度抓取方法 |
Country Status (1)
| Country | Link |
|---|---|
| WO (1) | WO2024113216A1 (zh) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107618030A (zh) * | 2016-07-16 | 2018-01-23 | 深圳市得意自动化科技有限公司 | 基于视觉的机器人动态跟踪抓取方法及*** |
| CN111775154A (zh) * | 2020-07-20 | 2020-10-16 | 广东拓斯达科技股份有限公司 | 一种机器人视觉*** |
| CN113379849A (zh) * | 2021-06-10 | 2021-09-10 | 南开大学 | 基于深度相机的机器人自主识别智能抓取方法及*** |
| US20210308866A1 (en) * | 2020-04-03 | 2021-10-07 | Fanuc Corporation | Adaptive grasp planning for bin picking |
| CN114299116A (zh) * | 2021-12-29 | 2022-04-08 | 伯朗特机器人股份有限公司 | 动态目标抓取方法、设备和存储介质 |
| CN114463244A (zh) * | 2020-11-09 | 2022-05-10 | 南京理工大学 | 一种视觉机器人抓取***及其控制方法 |
-
2022
- 2022-11-30 WO PCT/CN2022/135360 patent/WO2024113216A1/zh unknown
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107618030A (zh) * | 2016-07-16 | 2018-01-23 | 深圳市得意自动化科技有限公司 | 基于视觉的机器人动态跟踪抓取方法及*** |
| US20210308866A1 (en) * | 2020-04-03 | 2021-10-07 | Fanuc Corporation | Adaptive grasp planning for bin picking |
| CN111775154A (zh) * | 2020-07-20 | 2020-10-16 | 广东拓斯达科技股份有限公司 | 一种机器人视觉*** |
| CN114463244A (zh) * | 2020-11-09 | 2022-05-10 | 南京理工大学 | 一种视觉机器人抓取***及其控制方法 |
| CN113379849A (zh) * | 2021-06-10 | 2021-09-10 | 南开大学 | 基于深度相机的机器人自主识别智能抓取方法及*** |
| CN114299116A (zh) * | 2021-12-29 | 2022-04-08 | 伯朗特机器人股份有限公司 | 动态目标抓取方法、设备和存储介质 |
Non-Patent Citations (1)
| Title |
|---|
| XU-LIANG LIU: "Research on multi arm adaptive cooperative control method of intelligent manufacturing robot", MANUFACTURING AUTOMATION, vol. 44, no. 1, 25 January 2022 (2022-01-25), pages 110 - 113, XP093177068 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11562552B2 (en) | Detecting boxes | |
| CN110531707B (zh) | Scara机器人的摩擦模型改进以及动力学参数辨识方法 | |
| CN110238849B (zh) | 机器人手眼标定方法及装置 | |
| CN104552341B (zh) | 移动工业机器人单点多视角挂表位姿误差检测方法 | |
| CN111243017A (zh) | 基于3d视觉的智能机器人抓取方法 | |
| CN111267105A (zh) | 一种六关节机器人的动力学参数辨识及碰撞检测方法 | |
| CN110480615A (zh) | 机器人拆垛定位修正方法 | |
| CN104808490A (zh) | 一种面向模具保护的基于回声状态网络估计图像雅克比矩阵的无标定视觉伺服控制方法 | |
| CN112621746A (zh) | 带死区的pid控制方法及机械臂视觉伺服抓取*** | |
| CN114942593A (zh) | 一种基于干扰观测器补偿的机械臂自适应滑模控制方法 | |
| CN114378827B (zh) | 一种基于移动机械臂整体控制的动态目标跟踪抓取方法 | |
| Heyn et al. | Application of artificial neural networks in force-controlled automated assembly of complex shaped deformable components | |
| WO2024113216A1 (zh) | 一种工业模具智能制造机器人高精度抓取方法 | |
| CN112099505B (zh) | 一种低复杂度的移动机器人视觉伺服编队控制方法 | |
| Luo et al. | Robotic conveyor tracking with dynamic object fetching for industrial automation | |
| JPH0588721A (ja) | 関節型ロボツトの制御装置 | |
| CN111604900B (zh) | 一种适用于机械手在手操作的目标参数检测方法 | |
| CN108107882A (zh) | 基于光学运动跟踪的服务机器人自动标定与检测*** | |
| CN107443369A (zh) | 一种基于视觉测量模型逆辨识的机器臂无标定伺服控制方法 | |
| CN112123329A (zh) | 一种机器人3d视觉手眼标定方法 | |
| Song et al. | Automatic docking system of fuel filler with CAD model-based tracking and visual servoing control | |
| Qin et al. | Sensor calibration and trajectory planning in 3D vision-guided robots | |
| CN110940351A (zh) | 一种基于参数降维辨识的机器人精度补偿方法 | |
| CN108406773A (zh) | 一种耗能最少的2r欠驱动平面机械臂控制方法 | |
| Yong | RESEARCH ON PATH RECOGNITION OF WELDING MANIPULATOR BASED ON AUTOMATIC CONTROL ALGORITHM. |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 22966809 Country of ref document: EP Kind code of ref document: A1 |