A structure optimal design of training devices for the under-actuated hand rehabilitation

A method for the optimization of mechanical structure of the under-actuated hand rehabilitation training system is proposed to expand the motion range of its finger joints and to satisfy training requirements and it is based on analyzing spatial states of kinematic models. Simplified spatial models for the mechanical components and fingers are established. All possible spatial vectors under some mechanical size and finger length are calculated to form a set of spatial states by regarding the rotation angles of key mechanical components and finger joints as the elements of spatial vectors. Then, the optimal mechanical sizes are determined by investigating different state sets to make the motion range of finger joints meet the needs of clinical training. A software is used to build a 3D entity model for kinematic analysis. The results from simulation and prototype wear tests show that the proposed device can assist fingers to finish flexion and extension within required motion ranges and is capable of adapting the changes in finger length. It can be concluded that the proposed method is helpful for the structure optimal design of under-actuated limb rehabilitation training devices.