Abstract
Different predictive control strategies have been validated on a high-speed flexible
manipulator. It has been shown that the use of a rigid model instead of a flexible two-mass-spring model
could enhance the tuning procedure of Predictive Functional Control, and that the use of adequate
trajectory planning is essential to ensure the cancellation of vibrations. Cascaded predictive control,
which consists of a cascaded loop where the traditional servo algorithms are replaced by PFCs, could
minimize the peak vibrations and dramatically enhance the cycle time. Predictive Functional Control
alone is simpler to tune and can exhibit comparable performances, except that the controller is more
sensitive to nonlinear phenomena such as dry friction, which were not taken into account into the model
and generate a static error.
