Comparison Of Advanced Control Techniques For Motion Intention Recognition Using Emg Signals

This paper presents a comparative study between advanced neural networks and fuzzy logic techniques for recognition of cylindrical, spherical and strength griping types of the hand, by means of the analysis of acquired surface signals using a SHIELD-EKG-EMG card. The main motivation of this article is to show which technique is more efficient for the recognition of the three types of grip due to the versatility of uses in the hand using low cost hardware. The training stage consisted on a total of 147 training samples distributed in 49 training data for each type of study grip. In the experimental phase, 200 samples were acquired from a group of 10 participants and distributed in 10 samples in 2 sets of test for each type of grip. The results show that the difference in the recognition stage is not significant, so more advanced signal processing techniques will be put into effect in further studies to determine which technique is better.