| Summary: | 碩士 === 大同大學 === 機械工程研究所 === 89 === This paper focuses mostly on the controller design of the parallel robot and applications of virtual reality. The PID controller has been chosen as a wide usage in industries. But the adjustment of gain value is an inefficient and a troubling procedure. Therefore it needs to depend on a professional who has enough experience. This paper shows the usage of the back-propagation neural network to find the best gain value for the PID controller, and uses it to control the complicated hardware structure of the parallel robot. Through experiments, this control system can calculate a possible good value adjustment in the short time.
Main purpose of virtual reality (VR) is the interaction between operator and robot. Through human wisdom and the haptic device, one can real-time control a robot accomplishing complicated, heavy or high risking jobs. In the structure of the haptic device, there is visual and touch equipment. This study presents, via touch equipment, the human arm can directly control the parallel robot. And after the plan of the computer program, the operator can feel the change of the VR environment. Though the real-time show in the screen, one can see the change of the VR environment during operating process. In the experiment, we build the virtual spring and damper models, and simulate the process of the cutting virtual clay. Then the operator can feel really the existence of the VR environment.
|
|---|
