Petri Net Based Task Planning Technology for an Intelligent Underwater Robotic Manipulator System

• Main Authors: Shih Liang Chen, 陳思亮
• Other Authors: Mu Der Jeng
• Format: Others
• Language: en_US
• Published: 2000
• Online Access: http://ndltd.ncl.edu.tw/handle/05043420826869625147

碩士 === 國立海洋大學 === 電機工程學系 === 88 === The underwater environment is complex, uncertain, and dangerous. Therefore, using a robotic manipulator system to carry out underwater operations is indeed necessary. In view of the low autonomy and the necessity of total manual control in traditional robotic manipulators and in order to increase the manipulation convenience and efficiency, this thesis proposes a task planning technology for an underwater robotic manipulator system. We applied Colored Petri Nets to define the task model for underwater operations. This method aims at disassembly processes, and it can automatically generate the task planning sequences, and cope with errors using automatic recovery schemes. Following the above method, we design a tele-assisted-autonomous and Petri-net-based automatic planner. It has good man-machine interface and PC Windows execution environment. By using the mouse under the Windows, we can click & drag to define the definitions of objects, object destinations, and partial object precedence orders. Finally, we apply this task planner to an underwater robotic manipulator implemented by ourselves, and then test and verify the correctness and the feasibility of the proposed technique via this system.