logo SBA

ETD

Archivio digitale delle tesi discusse presso l’Università di Pisa

Tesi etd-11082017-133632


Tipo di tesi
Tesi di laurea magistrale
Autore
PELAEZ MURCIEGO, LUIS
URN
etd-11082017-133632
Titolo
Simultaneous and proportional EMG finger control for a rehabilitation exoskeleton hand: a synergy-based approach.
Dipartimento
INGEGNERIA DELL'INFORMAZIONE
Corso di studi
EMBEDDED COMPUTING SYSTEMS
Relatori
relatore Prof. Frisoli, Antonio
correlatore Dott. Barsotti, Michele
Parole chiave
  • bayes
  • bayesian
  • classification
  • classifier
  • control
  • controllo
  • dextrous
  • EMG
  • exoskeleton
  • finger
  • hand
  • mano
  • myoelectric
  • NMF
  • prosthetic
  • protesi
  • robotic
  • robotica
  • synergy
Data inizio appello
24/11/2017
Consultabilità
Completa
Riassunto
In this pilot study an approach for simultaneous and proportional myoelectric control of individual fingers is developed. This method is based on the non negative matrix factorization (NMF) of the bayesian estimators of the forearm EMG signals, which extract the synergies for each degree-of-freedom (DOF) using a semi-supervised algorithm with a short training phase. A synergy-based classifier is then introduced to select only the control signals corresponding to the active DOF’s. The first experiment is based on the second iteration of the NinaPro database, which includes the force at the fingertips of 40 different subjects. A second experimental protocol is proposed by using an exoskeleton hand specifically developed for stroke patient rehabilitation. The promising results show this method has the potential of providing a dexterous hand control for clinic rehabilitation and prosthetic control.
File