Paper
Neural-Genetic Model for Muscle Force EstimationBased on EMG Signal
- Authors:
- Caihua Xiong; Usama J. Naeem
- Abstract
- Theaim of this work is to propose a new model to carry out muscle force from electromyography (EMG) signal. Our neural-genetic model consists of two stages (genetic and neural) to predict human muscle force from the right arm muscles;where, the first (genetic stage) is employed to transform the raw EMG signal into muscle activation. The second (neural stage) uses Back-Propagation Neural Network (BPNN) method to extract muscle force from the muscle activation of the first stage. The neural-genetic model can efficiently extract muscle force features from raw EMG signals without passing the signal through signal processing steps. Our results showed that the regression of our neural-genetic model exceeded 99%. We used the mean square error (MSE) to measure the performance of our model. The MSE result of our neural-genetic model was very small.
- Keywords
- EMG Signal; Muscle Force; Genetic Algorithm; Neural Network
- StartPage
- 301
- EndPage
- 309
- Doi