| Summary: | 碩士 === 國立臺灣科技大學 === 機械工程系 === 102 === Programming an industrial robot by the typical teaching method, through the use of the robot teach pendant, is a tedious and time-consuming task that requires technical expertise. This thesis presents, 3D point cloud data based robotized spray painting system to generate robot motion plan and subsequently control the process in a much flexible and efficient way than the traditional teaching method. A graphical user interface is developed to define system’s input parameters, visualize the model and inspect the end-effector path for safety. Generally the system comprises of a KinectFusion based 3D scanning subsystem, a 6-axis painting robot, robot feeder subsystem and path planning and robot program generation unit. Moreover, we investigated various end-effector trajectories generation methods based on various robot systems and robot feeder subsystems. The trajectory generation has been achieved is in two modes: by extracting the curve-skeleton of the workpiece's 3D model or manually picking a number of critical corner points graphically. This research successfully implemented the two modes of operation of spray painting applications on YASKAWA MOTOMAN EPX2050 painting robot in integration with two conveying systems by using reconstructed 3D point cloud data of various bicycle-frames.
|
|---|
