Robust throwing design based on dynamic sensitivity analysis
Masafumi OKADA, Shota ONIWA, Wataru Hijikata
- Year
- 2018
- Citations
- 4
- Access
- Open access
Abstract
Throwing an object by a powered robotic system is an effective way for object manipulation in long distance. The focus of the throwing is on the accuracy of the landing point with respect to model uncertainties or disturbances. A robust controller is often designed, however, the motion will be finished before the controller produces its effects because throwing is fast and highly dynamic. Moreover, the robot system sometimes has zero adjustment error in the initial position on its joint angle. This error cannot be overcome by a robust controller. So far, we have proposed a dynamic sensitivity analysis method of throwing for a feed-forward controlled manipulator. The sensitivity of the landing point with respect to zero adjustment error has been calculated, and robust throwing with small sensitivity has been designed. In this paper, the conventional method is applied to a feed-forward/back controlled manipulator in the real world, and evaluations are executed by using a prototyped three-link manipulator. The effectiveness of the proposed method is evaluated based on the sensitivity and variance of the landing point, and a robust throwing with small sensitivity is designed.
Keywords
Related papers
Statistical Learning Theory
Yuhai Wu, Vladimir Vapnik
1999
Artificial intelligence: a modern approach
1995
Applied Nonlinear Control
Jean-Jacques Slotine, Weiping Li
1991
A new optimizer using particle swarm theory
R.C. Eberhart, James Kennedy
2002