Open AccessGait Phase Classification Based on sEMG Signals Using Long Short-Term Memory for Lower Limb Exoskeleton Robot
Author(s) -
Ye Yuan,
Ziming Guo,
Can Wang,
Shengcai Duan,
Lufeng Zhang,
Xinyu Wu
Publication year - 2020
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/853/1/012041
Subject(s) - exoskeleton , robustness (evolution) , gait , robot , computer science , gait analysis , gait cycle , ground reaction force , artificial intelligence , simulation , physical medicine and rehabilitation , kinematics , medicine , biochemistry , chemistry , physics , classical mechanics , gene
In this work, we present a Long Short-Term Memory Model (LSTMM) for gait phase classification based on sEMG signals to control the lower limb exoskeleton robot which can recognize 2 phases (Stand and Swing) of leg phases between the foot and ground. This model only needs four sEMG signals to control the lower limb exoskeleton robot helping the hemiplegia patient walking. Compared with the existing methods, the proposed model not only avoids the complex sensor systems but also enhances the accuracy of gait phase classification. The experimental results first verify the availability of sEMG data acquisition system on the lower limb exoskeleton robot made by the Shenzhen Institutes of Advanced Technologies (SIAT) by quantify the system with gold standard optoelectronic system Vicon, then show that the proposed LSTMM is significantly higher on prediction accuracy and has better robustness for gait phase classification to control the lower limb exoskeleton robot with different speeds. Finally, the maximum accuracy of LSTMM on the gait phase classification is 97.89%.