Estimation of human ankle impedance during walking using the perturberator robot
Elliott J. Rouse, Levi J. Hargrove, Eric J. Perreault, Todd Kuiken
- Year
- 2012
- Citations
- 34
Abstract
Motivated by the ability to provide biologically inspired control systems to recently developed powered ankle prostheses, the impedance parameters of the ankle were determined during the stance phase of walking. Five subjects walked across a walkway that included a robotic platform to perturb the ankle. A 0.035 radian perturbation was randomly applied at four perturbation timing points in stance phase: 100, 225, 350 and 475 ms following heel strike. A parametric model was used to represent the impedance of the ankle in terms of stiffness, damping and inertial components. A novel method for removing the angle and torque profiles that resulted from walking was introduced that uses bootstrapping while taking the subtraction of multiple averaged perturbed and non-perturbed trials. The stiffness values were found to increase on average throughout stance phase, progressing from 1.0 Nm/rad/kg at the 100 ms timing point to 4.6 Nm/rad/kg at the 475 ms timing point. Likewise, the damping values increased during stance phase, progressing from 0.012 Nms/rad/kg at the 100 ms timing point to 0.038 Nms/rad/kg at the 475 ms timing point. Variance accounted for (VAF) was used to quantify agreement of the model with the experimental results and was consistently very high, averaging above 98% for all perturbation timing points. This work can be used to guide the development of biologically inspired powered ankle control systems by providing realistic impedance parameters of the ankle joint.
Keywords
Related papers
Statistical Learning Theory
Yuhai Wu, Vladimir Vapnik
1999
Artificial intelligence: a modern approach
1995
Fractional Differential Equations
Igor Podlubný
2025
Applied Nonlinear Control
Jean-Jacques Slotine, Weiping Li
1991