Human Arm’s Muscle Force Estimation Model Based on Fuzzy Theory

Abstract

Muscle force estimation, is important in order to understand human control strategies, rehabilitation, motion analysis and human-machinery interfacing. To estimate the force, bio-signals analysis is used. Bio-signals enable signal recognition and analysis of muscle action and it is used to estimate muscle force. In this work, a method to estimate human muscle force is proposed. This model, which is based on Fuzzy Logic Theorem, is fast and easy to implement. Fuzzy Model process a raw of rectified smoothed electromyography (RSEMG) signal and extract muscle force. Our model is a general one, it can use for any muscles. To prove that, we used biceps brachii muscle of the left and right arms and two different arm muscles as validation. Our results, demonstrate that the new model improves the accuracy of muscle force estimation. This model can efficiently extract muscle force features from (EMG) signals. Our results showed that algorithmic regression exceeds 99%. The mean square error (MSE) results for Fuzzy Model was very small (MSE equal to 1.11x10-08).