Tesi etd-12152017-145001
Link copiato negli appunti
Tipo di tesi
Perfezionamento
Autore
TARANTINO, SERGIO
URN
etd-12152017-145001
Titolo
The myokinetic approach for hand prosthetic control
Settore scientifico disciplinare
ING-IND/34
Corso di studi
INGEGNERIA - Biorobotics
Commissione
relatore Prof. CIPRIANI, CHRISTIAN
Parole chiave
- dexterous hand prosthesis control
- magnetic sources tracking
- optimization algorithm
Data inizio appello
30/06/2018;
Disponibilità
completa
Riassunto analitico
An upper limb amputation deprives individuals of their innate ability to manipulate objects. Such disability can be restored with a robotic prosthesis linked to the brain by a human-machine interface (HMI) capable of decoding voluntary intentions, and sending motor commands to the prosthesis. Clinical or research HMIs rely on the interpretation of electrophysiological signals recorded from the muscles (i.e. the electromyogram, EMG). However, the quest for an HMI that allows for arbitrary and physiologically appropriate control of dexterous prostheses, is far from being completed. In this thesis, we propose a completely novel HMI that aims to track the muscles contractions with implanted permanent magnets (i.e. magnetic markers), by means of magnetic field sensors. We called this a myokinetic control interface.
In the Introduction of this thesis, we present an overview of the state of the art of the control strategies currently used in powered prostheses (based on EMG signals) and the state of the art of magnetic markers tracking methods.
In Chapter 1, we present a first implementation and characterization of a myokinetic interface, which demonstrates how this method could allow to directly multiple degrees of freedom of a hand prosthesis through the localization of a number of magnetic markers. In particular, we present the design features of the myokinetic interface, and the implementation of a localizer which exploits six 3-axis magnetic field sensors to localize four magnetic markers in an anatomically relevant workspace. We also present Finite Element (FE) modelling aimed at assessing the effects of external magnetic interferences (e.g. external magnets or ferromagnetic objects) and their attenuation using shields. Furthermore, we demonstrate the possibility to use the myokinetic interface to simultaneously and independently control online four degrees of freedom of a hand prosthesis.
In Chapter 2, we exploit FE simulations to optimize the performance of the myokinetic interface. In particular, we quantified the impact of (i) the number of magnetic markers to be tracked, (ii) the number and spatial configuration of magnetic field sensors used, (iii) the magnets-sensor distance and (iv) the magnets-magnets distance on the tracking accuracy, precision and computation time of the multi magnet localizer. Such simulations were also experimentally validated by replicating the simulated system using a physical platform.
In the Introduction of this thesis, we present an overview of the state of the art of the control strategies currently used in powered prostheses (based on EMG signals) and the state of the art of magnetic markers tracking methods.
In Chapter 1, we present a first implementation and characterization of a myokinetic interface, which demonstrates how this method could allow to directly multiple degrees of freedom of a hand prosthesis through the localization of a number of magnetic markers. In particular, we present the design features of the myokinetic interface, and the implementation of a localizer which exploits six 3-axis magnetic field sensors to localize four magnetic markers in an anatomically relevant workspace. We also present Finite Element (FE) modelling aimed at assessing the effects of external magnetic interferences (e.g. external magnets or ferromagnetic objects) and their attenuation using shields. Furthermore, we demonstrate the possibility to use the myokinetic interface to simultaneously and independently control online four degrees of freedom of a hand prosthesis.
In Chapter 2, we exploit FE simulations to optimize the performance of the myokinetic interface. In particular, we quantified the impact of (i) the number of magnetic markers to be tracked, (ii) the number and spatial configuration of magnetic field sensors used, (iii) the magnets-sensor distance and (iv) the magnets-magnets distance on the tracking accuracy, precision and computation time of the multi magnet localizer. Such simulations were also experimentally validated by replicating the simulated system using a physical platform.
File
| Nome file | Dimensione |
|---|---|
| Tarantin...s_sub.pdf | 26.34 Mb |
Contatta l'autore |
|