Upper limb exoskeleton control based on sliding mode control and feedback linearization

Exoskeletons have potential to improve human quality of life by relieving loads on the human musculoskeletal system or by helping motor rehabilitation. Controllers that were initially developed for industrial applications are also applied to control an exoskeleton, despite of major differences in requirements. Nevertheless, which controller has better performance for this specific application remains an open question. This paper presents a comparison between sliding mode control and feedback linearization control. The implementation of the sliding mode controller assumes a complete measurement of the biological system's dynamics. On the other hand, the feedback linearization method does not need any information from the biological system. Our results indicate that a model combining biological control and dynamics could improve several characteristics of robotic exoskeleton, such as energy consumption.