Motion intensity modeling and trajectory control of upper limb rehabilitation exoskeleton robot based on multi-modal information

Abstract The motion intensity of patient is significant for the trajectory control exoskeleton robot during rehabilitation, as it may have important influence on training effect and human–robot interaction. To design rehabilitation task according to situation patients, a novel method designed based perception model. signal heart rate are collected fused into multi-modal information input layer vector deep learning framework, which used interaction model system. A 6-degree freedom (DOF) upper limb previously implement test. parameters iteratively optimized by grouping experimental data, identification analyzed compared. average recognition accuracy proposed can reach up 99.0% in data set 95.7% test set, respectively. results show that neural network (DNN) improve performance interaction, possible further training.