Synthesis of full-body 3-D human gait using optimal control methods
Authors:
- 1 May 2016
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1560-1566
Abstract
In this paper we present a method that uses optimal control for offline human gait synthesis that does not depend on motion capture data or task-specific controllers. Our method is based on efficient simulation of rigid multibody systems and a direct multiple-shooting method to solve the underlying space-time optimization problem. We formulated different optimization criteria and synthesized gaits for a fullbody 3-D human model with 34 degrees of freedom and compared the resulting movements with human data. By combining different criteria we are able to improve the similarity of the synthesized motions with respect to recorded human motioncapture motions.Keywords
This publication has 34 references indexed in Scilit:
- Locomotion skills for simulated quadrupedsACM Transactions on Graphics, 2011
- Generalized biped walking controlACM Transactions on Graphics, 2010
- Optimizing walking controllers for uncertain inputs and environmentsACM Transactions on Graphics, 2010
- From human to humanoid locomotion—an inverse optimal control approachAutonomous Robots, 2009
- Using optimization to create self-stable human-like runningRobotica, 2009
- Simulating biped behaviors from human motion dataACM Transactions on Graphics, 2007
- Compliant leg behaviour explains basic dynamics of walking and runningProceedings Of The Royal Society B-Biological Sciences, 2006
- Efficient Factorization of the Joint-Space Inertia Matrix for Branched Kinematic TreesThe International Journal of Robotics Research, 2005
- Adjustments to Zatsiorsky-Seluyanov's segment inertia parametersJournal of Biomechanics, 1996
- Stabilization of Constrained Mechanical Systems with DAEs and Invariant ManifoldsMechanics of Structures and Machines, 1995