2026-09-21T11:26:31Zhttps://dspace.mit.edu/server/oai/requestoai:dspace.mit.edu:1721.1/838222022-01-13T07:54:07Zcom_1721.1_7582com_1721.1_7581col_1721.1_131022
Hugh M. Herr and Leonid A. Mirny.
Markowitz, Jared (Jared John)
Massachusetts Institute of Technology. Department of Physics.
Massachusetts Institute of Technology. Department of Physics
2014-01-09T19:58:51Z
2014-01-09T19:58:51Z
2013
http://hdl.handle.net/1721.1/83822
865575967
Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Physics, 2013.
Cataloged from PDF version of thesis.
Includes bibliographical references (pages 119-123).
In this thesis we present a data-driven neuromuscular model of human walking and its application to prosthesis control. The model is novel in that it leverages tendon elasticity to more accurately predict the metabolic consumption of walking than conventional models. Paired with a reflex-based neural drive the model has been applied in the control of a robotic ankle-foot prosthesis, producing speed adaptive behavior. Current neuromuscular models significantly overestimate the metabolic demands of walking. We believe this is because they do not adequately consider the role of elasticity; specifically the parameters that govern the force-length relations of tendons in these models are typically taken from published values determined from cadaver studies. To investigate this issue we first collected kinematic, kinetic, electromyographic (EMG), and metabolic data from five subjects walking at six different speeds. The kinematic and kinetic data were used to estimate muscle lengths, muscle moment arms, and joint moments while the EMG data were used to estimate muscle activations. For each subject we performed a kinematically clamped optimization, varying the parameters that govern the force-length curve of each tendon while simultaneously seeking to minimize metabolic cost and maximize agreement with the observed joint moments. We found a family of parameter sets that excel at both objectives, providing agreement with both the collected kinetic and metabolic data. This identification allows us to accurately predict the metabolic cost of walking as well as the force and state of individual muscles, lending insight into the roles and control objectives of different muscles throughout the gait cycle. This optimized muscle-tendon morphology was then applied with an optimized linear reflex architecture in the control of a powered ankle-foot prosthesis. Specifically, the model was fed the robot's angle and state and used to command output torque. Clinical trials were conducted that demonstrated speed adaptive behavior; commanded net work was seen to increase with walking speed. This result supports both the efficacy of the modeling approach and its potential utility in controlling life-like prosthetic limbs.
by Jared Markowitz.
Ph.D.
123 pages
eng
Massachusetts Institute of Technology
M.I.T. theses are protected by
copyright. They may be viewed from this source for any purpose, but
reproduction or distribution in any format is prohibited without written
permission. See provided URL for inquiries about permission.
http://dspace.mit.edu/handle/1721.1/7582
Physics.
A data-driven neuromuscular model of walking and its application to prosthesis control
Thesis
application/pdf
false
Publication
unknown
unknown
<Publication xmlns="https://www.openaire.eu/cerif-profile/1.1/" id="8bbaa488-23c7-49db-8861-cc78c2c41386">
<Type xmlns="https://www.openaire.eu/cerif-profile/vocab/COAR_Publication_Types">http://purl.org/coar/resource_type/c_1843</Type>
<Language>eng</Language>
<Title>A data-driven neuromuscular model of walking and its application to prosthesis control</Title>
<PublishedIn>
<Publication>
</Publication>
</PublishedIn>
<PublicationDate>2013</PublicationDate>
<Authors>
<Author>
<DisplayName>Markowitz, Jared (Jared John)</DisplayName>
<Affiliation>
<OrgUnit>
</OrgUnit>
</Affiliation>
</Author>
</Authors>
<Editors>
</Editors>
<Publishers>
<Publisher>
<DisplayName>Massachusetts Institute of Technology</DisplayName>
<OrgUnit />
</Publisher>
</Publishers>
<License>http://dspace.mit.edu/handle/1721.1/7582</License>
<Keyword>Physics.</Keyword>
<Abstract>In this thesis we present a data-driven neuromuscular model of human walking and its application to prosthesis control. The model is novel in that it leverages tendon elasticity to more accurately predict the metabolic consumption of walking than conventional models. Paired with a reflex-based neural drive the model has been applied in the control of a robotic ankle-foot prosthesis, producing speed adaptive behavior. Current neuromuscular models significantly overestimate the metabolic demands of walking. We believe this is because they do not adequately consider the role of elasticity; specifically the parameters that govern the force-length relations of tendons in these models are typically taken from published values determined from cadaver studies. To investigate this issue we first collected kinematic, kinetic, electromyographic (EMG), and metabolic data from five subjects walking at six different speeds. The kinematic and kinetic data were used to estimate muscle lengths, muscle moment arms, and joint moments while the EMG data were used to estimate muscle activations. For each subject we performed a kinematically clamped optimization, varying the parameters that govern the force-length curve of each tendon while simultaneously seeking to minimize metabolic cost and maximize agreement with the observed joint moments. We found a family of parameter sets that excel at both objectives, providing agreement with both the collected kinetic and metabolic data. This identification allows us to accurately predict the metabolic cost of walking as well as the force and state of individual muscles, lending insight into the roles and control objectives of different muscles throughout the gait cycle. This optimized muscle-tendon morphology was then applied with an optimized linear reflex architecture in the control of a powered ankle-foot prosthesis. Specifically, the model was fed the robot's angle and state and used to command output torque. Clinical trials were conducted that demonstrated speed adaptive behavior; commanded net work was seen to increase with walking speed. This result supports both the efficacy of the modeling approach and its potential utility in controlling life-like prosthetic limbs.</Abstract>
<Access xmlns="http://purl.org/coar/access_right"
>
</Access>
</Publication>