A visual anti-collision control method for an industrial robot arm

An industrial robot, visual anti-collision technology, applied in manipulators, program-controlled manipulators, manufacturing tools, etc., can solve problems such as limited vision of manipulators, errors and safety accidents

CN108247637BActive Publication Date: 2020-11-24CENT SOUTH UNIV
6 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Publication Date
2020-11-24

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
Patent Text Reader

Abstract

The invention discloses a visual anti-collision control method for an industrial robot arm. The visual anti-collision control method for the industrial robot arm comprises the following steps of 1, establishing a three-dimensional system of coordinates of the working space where the mechanical arm is located; 2, establishing an Elman network based recognition model of a to-be-grabbed object target; 3, establishing an optimal mechanical arm grabbing path model based on a multinuclear support vector machine; 4, collecting the target image of the to-be-grabbed object in real time and identifyingobject category; and 5, determining the gravity center of the object based on the object category, obtaining the grabbing terminal coordinates, obtaining the optimal mechanical arm grabbing path, anddriving the mechanical arm to grab the object. By means of the method, an ideal trace that no collision is caused, the dynamic characteristic meets the margin requirement, and the trace length and movement time are short is figured out in a calculating mode for the mechanical arm to grab the target object, the production efficiency is greatly improved, meanwhile, the personnel cost is reduced, andbenefits are brought to factories.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The invention belongs to the field of mechanical arm control computing, and particularly relates to a visual collision avoidance control method for an industrial robot arm. Background technique

[0002] With the advent of "Industry 4.0", the application of robotic arms in factories has become more and more extensive, and the efficiency and safety of traditional working methods of man-machine interaction in complex environments have received attention. The existing manipulator can be controlled by the teach pendant and the host computer. The teach pendant control is to control the joints of the manipulator through the joystick and buttons on the teach pendant, and then store the program after completing the predetermined working condition; the host computer controls It is programmed in the host computer and can complete the cycle of the established working conditions. The above two control methods of manipulators can only complete the cycle of fixed wo...

Examples

Embodiment Construction

[0066] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0067] like figure 1 As shown, a visual collision avoidance control method for an industrial robot arm includes the following steps:

[0068] Step 1: Build the 3D coordinate system of the workspace where the robotic arm is located;

[0069] Use the image acquisition area to cover the entire workspace of the robotic arm to grab the target object. The midpoint of the left and right cameras of the binocular ZED camera is used as the origin, and the line connecting the center of the right camera of the binocular ZED to the center of the left camera of the binocular ZED is the positive y-axis. The three-dimensional coordinate system of the working space of the manipulator is established according to the right-hand rule;

[0070] Step 2: Build a target recognition model for objects to be grasped based on Elman neural network;

[0071] Taking the accumulated ...