Modular neural networks applied to surface EMG signals for limb function identification

This paper describes the application of neural networks to limb function identification using a single channel surface EMG signal and discusses improvements made in the decision algorithm which not only simplifies the control system and reduces identification time but also provides a more continuous identification of limb functions during real time application. Simulation results show that the method takes only a few milliseconds to identify limb functions and has a good identification rate.