Show simple item record

dc.contributor.advisor O'Malley, Marcia K
dc.creatorPehlivan, Ali Utku
dc.date.accessioned 2017-08-07T18:17:54Z
dc.date.available 2017-08-07T18:17:54Z
dc.date.created 2016-05
dc.date.issued 2016-04-26
dc.date.submitted May 2016
dc.identifier.citation Pehlivan, Ali Utku. "Subject Adaptive Control Paradigms for Robotic Rehabilitation." (2016) PhD diss., Rice University. http://hdl.handle.net/1911/96617.
dc.identifier.urihttp://hdl.handle.net/1911/96617
dc.description.abstract As the majority of the activities of daily living involve distal upper extremity movement, eff ective rehabilitation of the upper limbs, especially the distal joints, is crucial. Due to their inherent capabilities to deliver intensive and repetitive therapy, robotic devices are increasingly being used for the rehabilitation of neurologically impaired individuals. However, not every robotic device or therapy protocol has been shown to promote plasticity-mediated recovery. It is necessary that the robotic therapy must be capable of engaging the participant. Furthermore, the mechanical design of the robotic device must exhibit specifi c properties, such as low apparent inertia and friction, isotropic dynamic characteristics, and minimal backlash, to support sophisticated interaction modes. In this thesis a subject adaptive controller, capable of adaptively estimating position-dependent subject input and providing only the required amount of assistance is presented. This controller aims to maximize the participants' engagement in their therapy. Features of the controller were validated via simulations and experiments, and clinical validation was conducted with an elbow-forearm-wrist exoskeleton, the MAHI Exo-II. Results highlighted limitations in both the hardware's workspace and in the controller's performance. To address this limitations a novel wrist-forearm exoskeleton, the RiceWrsit-S, is proposed and an improved minimally assistive (mAAN) controller is presented. The controller is capable of estimating subject input as a function of time, hence it can estimate subject input regardless of position dependency, as opposed to the subject adaptive controller proposed in the rst part of the thesis. Novel features of the controller algorithm for maintaining subject engagement via performance based challenge modulation while still satisfying ultimately bounded error performance are presented. The mAAN controller and consistency of the accompanying algorithms are demonstrated experimentally with healthy subjects and with one subject with incomplete spinal cord injury in the RiceWrist-S. The proposed controllers and the novel exoskeletal device provide a means for a more eff ective robot-aided rehabilitation of neurologically impaired individuals.
dc.format.mimetype application/pdf
dc.language.iso eng
dc.subjectRobotics
Nonlinear Control
Exoskeletal Devices
Robotic Rehabilitation
dc.title Subject Adaptive Control Paradigms for Robotic Rehabilitation
dc.date.updated 2017-08-07T18:17:54Z
dc.type.genre Thesis
dc.type.material Text
thesis.degree.department Mechanical Engineering
thesis.degree.discipline Engineering
thesis.degree.grantor Rice University
thesis.degree.level Doctoral
thesis.degree.name Doctor of Philosophy


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record