Study on exoskeletal rehabilitation robot for upper limb and its control method

A 5-DOF exoskeletal rehabilitation robot, which can implement single joint and multi-joint complex motions and provide ADL training for hemiplegic patients, is presented. In general, hemiplegic patients are unilaterally impaired, so the surface electromyogram (sEMG) signal in the healthy limbs can be extracted to drive the rehabilitation robot to assist patients' impaired limb to carry out rehabilitation exercises. Herein two methods were involved: IAV and AR were used to extract features of sEMG acquired from four upper limb muscles which contribute to focused activities of that upper limb. The extracted features were used as the input to a back propagation neural network (BPN) in the Levenberg-Marquardt (LM) algorithm, then a relationship was formulated between sMEG and the rehablilitation motion with six upper limb rehabilitation exercise motions as outputs. Experiments prove the effectiveness of this method, which is useful for patient to train the nervous system, improve blood circulation and keep a sense of proper motion and improve range of motion.