LLE-ELM Recognition Method for the Joint Range of Shoulder and Elbow Motion

Abstract

This paper presents an LLE-ELM recognition model, which can identify upper limb motions with EMG signal acquisition system. In this study, the recognizing target set is established by the different angles of the two joint motions, surface electromyogram (sEMG) of shoulder and elbow actions are used as information sources, and the time-domain feature of the sEMG is formed to the characteristic vector space. LLE is used to reduce the difficulty of data calculation, and ELM is used to improve the speed and accuracy of motion recognition. Result shows that the validity of the LLE-ELM mixed model to identify the joint range of motions is verified.