Robust 2½D Visual Servoing of A Cable-Driven Parallel Robot Thanks to Trajectory Tracking
Zane Zaķe, François Chaumette, Nicolò Pedemonte, Stéphane Caro
Abstract
Cable-driven parallel robots (CDPRs) are a kind of parallel robots that have cables instead of rigid links. Implementing vision-based control on CDPRs leads to a good final accuracy despite modeling errors and other perturbations in the system. However, unlike final accuracy, the trajectory to the goal can be affected by the perturbations in the system. This letter proposes the use of trajectory tracking to improve the robustness of 21/2D visual servoing control of CDPRs. Lyapunov stability analysis is performed and, as a result, a novel workspace, named control stability workspace, is defined. This workspace defines the set of moving-platform poses where the robot is able to execute its task while being stable. The improvement of robustness is clearly shown in experimental validation.