|
Two nonlinear approaches are presented in the input/output
sense for the stability of robot's compliant motion. The first
approach applied a modified fuzzy control theory and aimed at
controlling force and position in a non-confliciting way. The
controller which is based on self-tuning proportional-deveriate
(PD) controller achieved roboust deburring in unknow plants
that may be open-loop unstable. A conventional PD level
controller was retained but it's robustness was increased by
using the self-tuning factor to adapt parameters in which
controller to change plant conditions. The second approach used
small gain controller and aimed at the relationship defined
between the interaction forces and the manipulator end-position
traditionally. According to this approach, smaller complinace
either in a robot or in the environment lead to a narrower
stability range. We propose some methods to shift the desired
trajectory to the neighborhood of the environment's dege and
not violated the stability in this approach. Finally, by
knowing distubance- compliance function,environment dynamic and
robot dynamics with tracking controller,we can choose a
compensator to shape the contact force if the sufficient
stability condition is satisfied. Some experiments have been
performed to demonstrate that the proposed criterions we
proposed are perfromed quite well in the deburrning process.
|