Research And Development Of Real Time Control System Based On Machine Vision Of DELTA Robot

With the development of China’s economy and structure upgrading in industry areas, big changes has been undergoing in assembly line,where robots have replaced humans to do simple and repetitive works. In fields like food, medicine, electronics and other production industries, many works like packaging, testing and transferring are simple, labor-intensive and need clear and low-dust environments. So inertial small Delta parallel robots have been used to replace manual operation.Firstly, this article has retrieved the applications of Delta robot and the development and prospects of visual identification in the robot industry at home and abroad. This article also did some classifications on research dynamic of trajectory planning of robot and control technology.Secondly, according to the characteristics of the conveyor belt in packaging line, objects on the conveyor belt can be recognized and the center is extracted using imaging process technology. Calibrate the camera’s internal matrix then transform pixel coordinates into three-dimensional coordinates in the camera reference. Calibrate the transformation matrix between camera and robot, then we can transfer target coordinates into the robot reference and control the motion of Delta robot.In order to improve the Delta robot’s moving speed, a kind of time-optimal trajectory planning method is proposed.Divide the robot’s working area, select the center point of each area as standard point and convert them into the joint space using inverse kinematics. Construct joint space curve using 5th order B-spline interpolation. Fractional particle swarm optimal method is used to optimize the global optimum point and plan time-optimal motion curve. Improve the operating speed, under the premise of ensuring the joint angular velocity, angular acceleration, angular jerk smooth and other constraints. The fuzzy controller is used to divide the distribution of time node.According to the characteristics of the Delta robot’s repetitive actions, a variable gain repetitive learning control strategy is put forward to improve the precision of the Delta robot’s repetitive movements. The method adopts the derivative of hyperbolic tangent function to replace the original constant parameters, and introduces the error measured as a function of input parameters. This method can make the control trajectory converge to the given trajectory with high precision and faster convergence speed.Finally, this paper has built the robot control platform based on Labview. The control platform consists of the inverse kinematics of Delta robot and trajectory planning and motion control module. We use NI – 7842 R board to acquire and output signal, and control the servo motor to complete specified action.