FEATURE EXTRACTION AND CLASSIFIER IN THE DEVELOPMENT OF EXOSKELETON BASED ON EMG SIGNAL CONTROL: A REVIEW

Triwiyanto, Triwiyanto and Wisana, I Dewa Gede Hari and Mak'ruf, Muhammad Ridha (2020) FEATURE EXTRACTION AND CLASSIFIER IN THE DEVELOPMENT OF EXOSKELETON BASED ON EMG SIGNAL CONTROL: A REVIEW. Journal of critical reviews, 7 (12). pp. 879-885. ISSN 2394-5125

Text 197-1592557068.pdf Download (434kB)

Abstract

Exoskeleton has been widely developed for the purpose of assistive and rehabilitation. This study's objective is to evaluate exoskeleton design based on EMG signal. EMG signals can provide an overview of activity in muscles, moreover the limbs motion can be represented by EMG signals through the activity. Some researchers have developed an exoskeleton by utilizing the control process through EMG signals. The selection of the right feature extraction determines the success of the classifier. Therefore, in this study, the feature extraction used in exoskeleton development research is feature extraction in the time domain (TD) MAV, RMS, IEMG, WL, SSC, and ZC. Furthermore, the classifier often used to predict the motion of the exoskeleton is an artificial neural network based on multilayer perceptron with backpropagation, neural network based on fuzzy, and support vector machines, because it has better accuracy. Some exoskeleton development for future research is discussed at the end, which includes, control system, safety, and compensation.

Item Type:	Article
Uncontrolled Keywords:	EMG, Exoskeleton, Feature Extraction, Classifier
Subjects:	R Medicine > Medical Electronics > Biomedical Engineering R Medicine > Medical Electronics > Clinical Engineering R Medicine > Medical Electronics
Divisions:	Poltekkes Kemenkes Surabaya > Pusat Penelitan dan Pengabdian Masyarakat > Pubikasi
Depositing User:	Nanik Indra Putri Sari
Date Deposited:	24 Aug 2021 03:28
Last Modified:	24 Aug 2021 03:28
URI:	http://repo.poltekkesdepkes-sby.ac.id/id/eprint/2450

Actions (login required)

View Item