Rehabilitation robot control using the VSD method

Recently, rehabilitation robots attract much attention both from engineering and medical fields due to their accurate and repetitive training capabilities. Their automated training protocol reduces the loads on clinical staff and helps effective rehabilitation by monitoring the patient´s status in a quantitative manner from various sensor measurements. In the control of such robotic systems, back-drivable force control is essential for the user´s safety and active participation into the training. In addition to the robot´s back-drivability, a robotic guidance force algorithm is required to realize so-called `assist-as-needed´ rehabilitation scheme. In this research, a VSD (virtual spring damper) controller is implemented in a 3 DOF (degree of freedom) non-anthropomorphic robotic manipulator for its rehabilitation purpose. The VSD controller guides the user´s hand motion along a pre-defined trajectory while an admittance model allows the user to reflect his/her intention captured from a force sensor installed at the robot´s end-effector. The performance of the proposed control method is verified by experiments on a circular trajectory by varying the parameters of admittance model.