Design, fabrication, programming, and evaluation of an ankle foot simulator

Update Item Information
Title Design, fabrication, programming, and evaluation of an ankle foot simulator
Publication Type thesis
School or College College of Engineering
Department Mechanical Engineering
Author Miller, Jonathan James
Date 2016
Description Parkinson Disease (PD) is a progressive and chronic movement disorder that affects an individual's ability to walk and move naturally. Research shows that training using virtual reality can offer improvements over traditional therapy and decrease the effects of some PD symptoms. In an effort to address the need for such therapeutic intervention, a Virtual Reality (VR) rehabilitation simulator was developed using 3D graphical displays in concert with haptic Smart Shoes. The system creates challenging virtual terrain to safely train participants in situations that demand greater balance and neuromuscular control. As part of this effort, an Ankle Foot Simulator (AFS) was created to mimic human gait, including ankle and foot response to a variety of terrain features. This thesis describes the development and testing of a novel AFS robot designed to mimic human gait and evaluate Smart Shoe behavior and response to perturbations. The major design requirement for the AFS robot is to reproduce natural gait dynamics by: 1) matching complex trajectories of the ankle, 2) generating Ground Reaction Forces (GRF) during normal walking gait, and 3) mimicking foot/ankle dynamics such as ankle roll over. This thesis focuses on the design and control of the AFS to achieve sufficient Range of Motion (ROM) to mimic human gait, including extreme ankle rollover, while providing appropriately fast dynamics, sufficient load capacity, and high repeatability. Design aspects of the AFS include 1) forward and inverse kinematic derivations of the ankle mechanism, 2) derivations of feedforward components of the control algorithms, and 3) mapping ankle mechanism actuator forces to ankle moments. The AFS robot tracks ankle position and orientation data to within 5.5 mm and 5.5 degrees. The AFS is also able to reproduce GRFs, including dorsal/plantar flexion and inversion/eversion ankle moments with an r2 value of 0.82 or more. The AFS also demonstrates passive ankle stiffness. To improve performance of the AFS, an iterative learning controller is suggested for future work.
Type Text
Publisher University of Utah
Subject Awesome; Control; Design; Electro-Mechanical; Gait; Robot
Dissertation Name Master of Science
Language eng
Rights Management ©Jonathan James Miller
Format application/pdf
Format Medium application/pdf
Format Extent 5,216,570 bytes
Identifier etd3/id/4231
ARK ark:/87278/s6q84nf0
Setname ir_etd
ID 197776
Reference URL https://collections.lib.utah.edu/ark:/87278/s6q84nf0
Back to Search Results