Real time control of an upper limb orthosis robot for a passive rehabilitation

This study focuses on the design and the position control of a didactic robotic therapy system. This robot is an upper limb device emulating the passive therapy of a human elbow joint. Model Predictive controller (MPC), PI and RST controllers are designed to ensure good closed loop performances. A comparative study between these three controllers is established in a real time environment. This comparison highlights the effectiveness of the model predictive controller. The purpose of this robotic rehabilitation emulation test is to ensure smooth therapy without significant overshoot in order to assist therapists apply their rehabilitation program in good conditions. This test needs to be validated for clinic therapy.