Electrode Deviation Evaluation of sEMG during Wrist Motion Classification using SVM

With the escalating aging population, the number of individuals suffering from upper limb hemiplegia is steadily rising each year. The upper limb rehabilitation robot plays a vital role in solving the above problems. As an essential bioelectrical signal, Surface Electromyography (sEMG) is widely used in rehabilitation training. However, in practical applications, the deviation of sEMG electrodes often has a bad effect on the accuracy of signal acquisition. The focus of this paper is to delve into the impact of sEMG electrode deviation on signal acquisition and to propose a viable solution strategy. We collected the standard signals and the signals of electrode deviation, respectively, and fused the data. Specifically, we used the Support Vector Machine (SVM) algorithm to address the issues caused by electrode deviation. The experimental results show that the SVM algorithm effectively solves the errors caused by sEMG electrode deviation in classification problems.