Uncertainty Estimation and Performance Optimization for Vertical Articulated and Parallel Robot Manipulators

碩士 === 國立臺灣大學 === 機械工程學研究所 === 106 === Uncertainty in robot manipulations is a fundamental source of inaccuracy. This study estimates uncertainty in robot manipulations for both vertical articulated and parallel robot manipulators to improve accuracy. The dynamic models of manipulators with uncertai...

Full description

Bibliographic Details
Main Authors: Ying-Kuan Tsai, 蔡穎寬
Other Authors: Kuei-Yuan Chan
Format: Others
Language:zh-TW
Published: 2018
Online Access:http://ndltd.ncl.edu.tw/handle/4ftqxx
Description
Summary:碩士 === 國立臺灣大學 === 機械工程學研究所 === 106 === Uncertainty in robot manipulations is a fundamental source of inaccuracy. This study estimates uncertainty in robot manipulations for both vertical articulated and parallel robot manipulators to improve accuracy. The dynamic models of manipulators with uncertainty will be derived. Transmission error of harmonic drives were be considered in the vertical articulated robot, while joint clearance and joint wear were considered in the parallel robot. For vertical articulated robot manipulators, the values of uncertainty in harmonic drives of joints are identified through testing results of virtual experiment. After input these identified values of uncertainty parameters into the model, optimization of path parameters and workspace for a specific path can be proceeded. For the parallel robot, we make it operate under the same path for several times in computer simulation. It helps us to comprehend the changing of clearance values due to the increasing number of operation and how it influences the performance of robot manipulations. Similarly, optimization of path parameters and workspace was proceeded for the wear effect existing in parallel robots. This study applies the research procedure into the 3-degree-of-freedom (3-DOF) vertical articulated robot manipulator which is responsible for welding processes and the 5-bar parallel robot manipulator which is responsible for handling and picking tasks. The simulation results imply the accuracy of the welding processes can be improved at most by 51.06% and 38.85% for the single-objective and multi-objective optimization, respectively, while the accuracy of the handling and picking tasks can be improved by 31.05% via the quantification and estimation of ISO 9283 performance criteria.