Adaptive hybrid impedance control for a 3DOF upper limb rehabilitation robot using hybrid automata

There is a growing need for effective and adaptive robot-assisted rehabilitation platforms for post-stroke patients which can facilitate considerably their sensorimotor control performance, and also ensure safety for the patients. A 3-DOF adaptive robot-assisted rehabilitation platform is proposed in this work which uses at its core a hybrid impedance control framework to track simultaneously both desired force and position trajectory, while regulating the apparent impedance of the robot as seen by the patient to ensure robot-patient compliant motion. To make the system adaptive to the patient recovery process, the impedance characteristic of the patient's impaired limb is modeled as a parameter of recovery, and is estimated online using a recursive polynomial model estimator. A hybrid automata is then implemented to specify different apparent robot impedances to track the recovery process. Preliminary simulation results showed good response of the proposed framework to the changing patient's arm impedance profile as well as good trajectory tracking of force and position in task space.