Header logo is am

Neuromuscular Stochastic Optimal Control of a Tendon Driven Index Finger


Conference Paper


With the goal to build robotic hands which can reach the levels of dexterity and robustness of the hand, the question of what are the candidate control principles that can handle the nonlinearities, the high dimensionality and the internal noise of biomechanical structures of the complexity of the hand, is still open. In this work we present the first stochastic optimal feedback controller applied to a full tendon driven simulated robotic index finger. In our model we do take into account the full tendon structure of the index finger which consist of 11 tendons based on the underlying physiology and we consider muscle with the typical force - length and force velocity properties. Our feedback controller show robustness against noise and perturbation of the dynamics while it can also successfully handle the nonlinearities and high dimensionality of the robotic index finger. Furthermore as it is shown in the evaluations, it provides the complete time history of the tendon excursions and the tendon velocities of the index finger for the tasks of tapping with zero and nonzero terminal velocities.

Author(s): Theodorou, E. A. and Todorov, E. and Valero-Cuevas, F.
Book Title: Proceedings of American Control Conference (ACC)
Year: 2011

Department(s): Autonomous Motion
Bibtex Type: Conference Paper (inproceedings)
Paper Type: Conference

Cross Ref: p10306
Note: clmc

Links: PDF


  title = {Neuromuscular Stochastic Optimal Control of a Tendon Driven Index Finger },
  author = {Theodorou, E. A. and Todorov, E. and Valero-Cuevas, F.},
  booktitle = {Proceedings of American Control Conference (ACC)},
  year = {2011},
  note = {clmc},
  crossref = {p10306}