Estimation of elbow joint angle from Time domain features of SEMG signals using Fuzzy Logic for prosthtic control

Abstract

In this work an attempt is made to estimate the elbow joint angle from Surface Electromyography (SEMG) signal during dynamic contraction using Fuzzy logic technique. Here the SEMG signals are taken from the biceps brachii of subjects during flexion and extension of eblow. To estimate the elbow joint angle, the SEMG signals are segmented into 250ms by adjacent window technique and two time domain parameters like Integrated EMG (IEMG) and Zero crossing (ZC) are extracted from windowed Raw EMG signals. If-then rules of fuzzy logic are derived from the experimental findings. The estimated values of elbow joint angles are compared with the actual angle values. A two Dimensional robotic arm animations is also coded using LabVIEW and are incorporated to the output of fuzzy logic system to simulate the estimated angle. The system is validated using regression value. Regression value obtained from the experiment is 0.7975.