Estimation of Lower Limb Periodic Motions from sEMG Using Least Squares Support Vector Regression

Abstract

In this paper, a new technique for predicting human lower limb periodic motionsfrom multi-channel surface ElectroMyoGram (s EMG) was proposed on the basis of least-squares support vector regression (LS-SVR). The sEMG signals were sampled from sevenhuman lower limb muscles. Two channels sEMG were selected and mapped to muscle acti-vation levels for angles estimation based on cross-correlation analysis. To deal with the timedelay introduced by low-pass filtering of raw s EMG, a k-order dynamic model was derivedto represent the dynamic relationship between the joint angles and muscle activation levels.The dynamic model was built by data driven LS-SVR with radial basis function kernel. Theinputs of the LS-SVR are muscle activation levels, and the outputs are joint angles of the hipand knee. In experiments, 48 s EMG-angle datasets sampled from six healthy people wereutilized to verify the effectiveness of the proposed method. Result shows that the humanlower limb joint angles can be well estimated in different motion conditions.

Tasks

Reproductions