University of Twente Student Theses


Optimal Potential Shaping on SE(3) via Neural Approximators

Wotte, Y. P. (2021) Optimal Potential Shaping on SE(3) via Neural Approximators.

[img] PDF
Abstract:This work combines optimal control and energy balancing, passivity based control (EB-PBC) on the Lie group SE(3), which is the configuration space of rigid bodies. Generally, EB-PBC achieves stable interactions with unknown environments by explicitly keeping the energy of a closed-loop system bounded. In the case of rigid bodies on SE(3), this recently allowed deriving impedance control based on a quadratic energy. However, choosing such a quadratic control-law is not connected to any principles from optimal control, which makes it an arbitrary choice. The derivation is phrased as an optimal control problem to extend such geometric impedance control beyond the quadratic case. Neural Nets and the Lie Group structure of SE(3) are used to conveniently solve the arising non-trivial problem of optimization. The final algorithm is validated on a state-regulation task.
Item Type:Essay (Master)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:31 mathematics, 53 electrotechnology, 54 computer science
Programme:Electrical Engineering MSc (60353)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page