Back to search

National Technical University of Athens (NTUA)

Μεθοδολογίες προγραμματισμού κίνησης σε νευρο-ρομποτικά συστήματα

Abstract

dc:description

Conceptually, a robotic device may be interfaced with a human neural system. Such a robot processes electro myographic activity (EMG signals) and uses it in the control loop that moves the mechanical links. In realizing such a neuro-robotic system, several challenges need to be met. One such challenge is using noisy neural signals as reference inputs in the feedback loop; another is the requirement to mimic human behavior during obstacle avoidance and interaction with non-planar surfaces. It is difficult to use only human neural signals to control an artifact. A human arm cannot use the sensory feedback from the robot in order to directly control it, and the artifact does not have the exact same morphology as the human arm to be able to interpret the neural signals perfectly. This mismatch can jeopardize the person and inflict damage on the environment. Despite the significant amount of work on modeling and decoding the signals from the human brain or from a part of the human body, for safety reasons it is necessary to establish a low level control loop around the artifact, for the exoskeleton’s collision avoidance and motion control. Further motivation for the design of such a control loop comes from robotic tele-operation applications, where the robot’s environment may be dramatically different compared to that in which the human operates. Then, analysis of the human’s signals alone may be insufficient to safely complete the desired task. Another issue that needs to be considered in the design of neuro-robotic systems, (consisting of an exoskeleton cooperating with the human upper limb), is the fact that motors must apply ppropriate torque in order for the exoskeleton to aid, rather than impede, arm movement. The design of most exoskeletons is based on neuro-scientific models, which dictate an agonistic-antagonistic behavior of the mechanism joints. The robotic device motion is therefore constrained. In this thesis we address the challenge of designing such controllers for the artifact. Our controllers drive the robot with safety for surface manipulation and compliant motion without collisions with objects in the environment, while considering input constraints which reflect motion limitations. The consideration of the control design problem for a redundant manipulator, the joint rate inputs of which must remain within pre-specified bounds, is addressed. The development of globally uniformly asymptotically stable controllers is established for this type of robots, for the purpose of: • reference trajectory tracking with obstacle avoidance, and • stabilization with obstacle avoidance, on 2-D surfaces which are embedded manifolds in 3-D robot workspaces.

Degree

thesis:*
Grantor dc:publisher
National Technical University of Athens (NTUA)
Year dc:date
2009

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Papageorgiou, Xanthi
  • Παπαγεωργίου, Ξανθή

Subjects

dc:subject × 9

Rights

Language dc:language
gre

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/17788

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Papageorgiou, Xanthi; Παπαγεωργίου, Ξανθή. Μεθοδολογίες προγραμματισμού κίνησης σε νευρο-ρομποτικά συστήματα. National Technical University of Athens (NTUA), 2009. http://hdl.handle.net/10442/hedi/17788