Скачать или смотреть A dual controller/estimator approach for constraint-based task control (presentation)

This work demonstrates an approach that leverages eTaSL to not only perform control but also estimation for
sensor-based robot applications.

The advantage of this approach is demonstrated on a contour following application, where the robot has
to follow an unknown contour using (noisy) force/torque signals.

Exploiting the duality between estimation and control, we separately describe the estimation and control. While parameters
of estimation are determined by sensor dynamics and noise, the control parameters are determined by robot and contact dynamics.
Separately describing these and tuning them allows for higher performance, i.e. reduced tracking errors and increased contour
following speeds.


For more information, refer to the following publication:

R. Viljoen, J. Ubbink, G. Pipeleers, W. Decré, E. Aertbeliën and J. De Schutter, Enhancing Constraint-Based Robot Task Specification with Dual Controller and Estimator Synthesis, 2023 IEEE 19th International Conference on Automation Science and Engineering (CASE), Auckland, New Zealand, 2023, pp. 1-8, doi: 10.1109/CASE56687.2023.10260605.