Design of a controller for a rehabilitation robotic system with one degree of freedom

Post stroke impairments are one of the pathologies most demanding for rehabilitation. Regular rehabilitation is very expensive both economically and socially and it requires physiotherapists to work very hard. Robotic rehabilitation is seen as an excellent substitute to the regular rehabilitation. This work deals with the controller of the actuators of a generic exoskeleton system for the upper limb rehabilitation. The exoskeleton actuator control is performed via the Simulink software in combination with the CPU of a normal computer. The use of Simulink ensures a full, easy and straightforward customisation of the system response. So the patient/physiotherapist's needs can be fulfilled without subverting the system stability. The actuator control was tested with signals related to one person's muscle activity, simulating a rehabilitation session with an active role of the patient. Furthermore, many tests were performed with signals generated in Simulink and thought to be used as exercise on a passive patient. Hence, the designed controller showed to successfully control an exoskeleton actuator with great accuracy, repeatability and with a good timing. These results were confirmed both in patient in charge and robot in charge tasks.Post stroke impairments are one of the pathologies most demanding for rehabilitation. Regular rehabilitation is very expensive both economically and socially and it requires physiotherapists to work very hard. Robotic rehabilitation is seen as an excellent substitute to the regular rehabilitation. This work deals with the controller of the actuators of a generic exoskeleton system for the upper limb rehabilitation. The exoskeleton actuator control is performed via the Simulink software in combination with the CPU of a normal computer. The use of Simulink ensures a full, easy and straightforward customisation of the system response. So the patient/physiotherapist's needs can be fulfilled without subverting the system stability. The actuator control was tested with signals related to one person's muscle activity, simulating a rehabilitation session with an active role of the patient. Furthermore, many tests were performed with signals generated in Simulink and thought to be used as exercise on a passive patient. Hence, the designed controller showed to successfully control an exoskeleton actuator with great accuracy, repeatability and with a good timing. These results were confirmed both in patient in charge and robot in charge tasks