International Conference on Advances in Computer Science and Electronics Engineering - CSEE 2014
Author(s) : HASLINIBINTI ABDULLAH , TAKEHIKO OGAWA
Recently, the myoelectric potential generated by the movement of human muscle used as interface for controlling a robot arm and prosthetic device has been studied. The changes of myoelectric potential by muscle motion gave the impact of the robot arm control. It is proposed that electromyography (EMG) patterns can be analyzed and clarified by Neural Network (NN) for the motion determination. Thus, this paper proposes a pattern estimation method of forearm motion by Complex-Valued Neural Network (CVNN).The forearm motion was recorded by Surface Electromyography (SEMG) method as the analysis data from the area of the back and frontof the arm intermediate part. We focused on how to estimate between the four movements in this study using Complex-Valued Neural Network (CVNN). The results show how the methodology we adopted allows us to obtain good accuracy in estimating the forearm motion.