| Summary: | 碩士 === 國立成功大學 === 機械工程學系 === 88 === Most robot manipulators used in the industry today are specially programmed for the specific tasks they perform. To increase the flexibility of manipulators, it is desirable to utilize visual feedback information to adapt themselves to the changes from the environment. Active vision is a good way to imitate human visual behavior through a servomechanism and stereo cameras. The cameras are operated in an active environment that can be controlled to move, to take pictures at the desired positions and orientations, and to decide its actions with respect to the feedback of the observations.
The objective of this thesis is to propose a new visual servo controller and a novel coordinated visual guided control strategy for the binocular robotic head and robot manipulator. Fuzzy sliding mode control is applied to the proposed visual servo controller, which integrates both the kinematics and dynamics of the servomechanism. The proposed coordinated control strategy is based upon the two visual servo controllers for binocular head and manipulator. In addition, the computations of the optical flow using Sum of Squared Difference (SSD) method and Kalman filters for the prediction of target’s trajectory are required.
Finally, the theoretical results will be experimentally verified on the integrated system that includes a robotic binocular head and a robot manipulator. Specific experimental demonstrations will include active visual tracking of the binocular head and coordinated control of the binocular head and robot manipulator.
|
|---|
