| Summary: | 碩士 === 國立雲林科技大學 === 機械工程系 === 105 === This research investigates the Scara robot in rapid positioning. The load inertia is uncertain in advance. Using the same servo gains for all payloads causes some problems, like overshoot, positioning errors and lower efficiency. The research goal is to enable the system to identify the load inertia, and tune gains automatically.
In the thesis, an industrial Scara robot is used as the experimental platform. Firstly, we deduce and analyze the dynamic system. The dynamic equation is used to simulate the positioning effect and performance of the system. A table relating the load to the optimal controller gains is established by the response of the experiments, and is integrated into the controller. Finally T-Curve trajectories for the target are used to optimize the reaction. Experiments show that the settling time with auto-gain-tuning and T-Curve trajectories is reduced by more than 40% compared to the situations with
fixed gain control.
Keywords：Scara、robot arm、auto-tuning、load identification、T-Curve、pathplanning
|
|---|
