Fuzzy Logic Design of a Robotic Manipulator Combined with Computer Vision for Shape Sorting Motion Control

• Main Authors: Jia-Ching Lu, 呂家慶
• Other Authors: Jung-Shan Lin
• Format: Others
• Language: en_US
• Published: 2006
• Online Access: http://ndltd.ncl.edu.tw/handle/37888209702519592290

碩士 === 國立暨南國際大學 === 電機工程學系 === 94 === In this thesis, a fuzzy logic design scheme with computer vision is developed for the shape sorting motion control of a robotic manipulator system combined with a CCD camera. The 5-axis robot arm ``MENTOR" employs a CCD camera, set on the end point of the robot arm, in order to obtain the image information from workspace for achieving the purpose of visual tracking. The major control objective of the robot arm system is not only to distinguish various shapes of objects and sockets on the workspace, but also to pick up the correct object and place it into the desired destination by applying fuzzy feedback control for position accuracy. The theoretic analysis of robotic manipulator is introduced and investigated in the beginning, including system dynamics and change of coordinates. In addition, several image processing techniques are applied to obtain the appropriate image information for visual tracking. The purpose of using these techniques is for edge detection, angle rotation and position determination in order to assist the robot arm to achieve the objective of shape sorting motion control. For the sake of position precision, fuzzy logic design scheme is applied to be a key portion of our control strategy. The fuzzy rules with eight different control tactics are built up from two input sets of membership functions to eliminate the position error during the placing motion. Finally, a practical experiment of robot arm system with computer vision is built up to achieve the objective of shape sorting motion control, and a control GUI designed by Visual Basic is provided for the user-friendly purpose.