| The purpose of this dissertation research project is to use cosine tuning functions to extract position and velocity information from surface electromyographic signals (sEMG). These signals are obtained from the antagonistic muscle pairs located in the human arm. Such information could be used by others, for example, to control and teleoperate a humanoid robotic arm such that the robot arm moves and behaves in a similar manner corresponding to the human arm. Specifically, these sEMG signals are measured from a human subject's biceps and triceps muscles when the subject performs elbow flexion motions at various elbow angle positions. These motions are performed in the sagittal plane with a load applied to the subject's wrist. Thus, creating a platform for developing human-machine interface systems using electromyographic information is the focus of this project. Ultimately, we wish to use sEMG signals as a gauge for measuring a person's intention when they perform arm reaching motions and incorporate this information into an intelligent software agent for a biologically inspired, humanoid robot control architecture. |
