A Flatness-Based Model Predictive Control Strategy Using B-Splines Applied to a Gantry Robot

  • Klemens Springer (Speaker)

Activity: Talk or presentationContributed talkunknown

Description

In this contribution a model predictive control design utilizing the special properties of flat systems is presented. The concept is based on a parametrization of trajectories for the flat outputs of an arbitrary system using basis splines. Hence the computation of the required control inputs as well as predicted system states for the entire horizon is done without time integration. Through the variation of the order of the basis spline functions, the possibility of varying the continuity of the control input is given independent of the desired trajectory. In order to overcome difficulties caused by longer sampling times due to the computational burden induced by the optimization process, the concept is extended for combination with a cascaded, fast position control. Experimental results of the realization on a robotic system with fast dynamics demonstrating the expected behavior are presented.
Period11 Sept 2013
Event title22nd International Workshop on Robotics in Alpe-Adria-Danube Region
Event typeConference
LocationSloveniaShow on map

Fields of science

  • 202035 Robotics
  • 203022 Technical mechanics
  • 203013 Mechanical engineering
  • 203015 Mechatronics

JKU Focus areas

  • Mechatronics and Information Processing