Hand Exoskeleton System And Its Control Technology

The hand exoskeleton system provide an new way for the hand rehabilitation after stroke. As the hand is very dexterous and tasks of hand are complicated, it’s very important to study in the key technologies of hand exoskeleton, including structure design, motion control, interaction force control and application.There are 22 DOFs (degree of freedoms) in human hand, which is hard to be designed in the hand exoskeleton. This paper presents a hand exoskeleton with 3 fingers and 10 DOFs after analyzing the anatomy of human hand. According to the structure, we obtain the kinematics model as well as the dynamics model.The interaction between human and exoskeleton is nonlinear and time-varying in the practical application. To solve these problems, the interaction impedance is modeled, whose parameters are identified online via the forgetting factor recursive least square method. Thus we obtain the perturbation range and nominal values of the parameters through experiments in different postures. For the hand exoskeleton controlling, the motion control method is given out according to the dynamics model of hand exoskeleton, while the force control method, impedance control, is presented. That’s the basis for the hand exoskeleton interaction force control. In order to solve the nonlinear and time-varying characteristic of human-machine interaction, we present two different methods, robust control and model prediction reference adaptive control. One is taking the control plant as an uncertain model and design^.-synthesis robust controller to solve it. The other one is design model prediction reference adaptive controller. Both of them are proved effective in series of experiments. We also find that the robust controller could adjust the system back stable faster, while the adaptive controller makes the processing of adjusting much gentler, which meets the mechanical impedance of human-machine interaction.During the rehabilitation after stroke, not only the muscles but also the brain plasticity needs to be trained. This paper presents combining the electroencephalogram (EMG) signal processing technology and hand exoskeleton control technology to improve the rehabilitation effect. When the patient imagines the hand grasp movement, the EEG signals are processed by the fast independent component analyze method (FastICA) to obtain the patient’s grasp intention. Sending the intention command to the hand exoskeleton control system, the hand exoskeleton could move under the patient’s control. The results of experiments show the effectiveness of the application method.This paper presents a design method of hand exoskeleton system, as well as its control technologies. Besides, an application technology combining EEG and hand exoskeleton is presented. These support the further research in hand exoskeleton and its application.