Controlling Robot Hand Using FFT as Input to the NN Algorithm. (July 2019)

• Record Type: Journal Article
• Authors: Andrean, Deni; Pamungkas, Daniel S; Risandriya, Sumantri Kurniawan
• Is Part Of: Journal of physics. Volume 1230(2019)
• Publication Date: 2019-07
• Subjects: Physics -- Congresses; 530.5
• DOI: 10.1088/1742-6596/1230/1/012030
• Languages: English
• ISSNs: 1742-6588

Abstract

Abstract: To control the prosthetic hand can be used several methods; one of them is by training the system to know the movement of the muscle. This method is using an EMG sensor to read the frequency of the movement of the muscle. In this study, Myo Arm Sensors is used as a sensor and used a Neural Network algorithm. The frequencies of the signal from the sensors are measured when the hand of the subject is open, grip or half open. Moreover, these signals are grouped using octave band methods. This information is used to be learned by the system. The system enables to mimic the movement of the subject up to 87.5% accuracy.