Design and Fuzzy Sliding Mode Admittance Control of a Soft Wearable Exoskeleton for Elbow Rehabilitation

Exoskeletons composed of rigid frames and joints have been widely implemented in rehabilitation application. However, the rigid mechanical structure introduces many problems, such as heavy weight, large inertia, and joint misalignment. To overcome these undesirable inherent disadvantages of rigid exoskeletons, the innovative contributions of this paper are to investigate the design and fuzzy sliding mode admittance (FSMCA) control of a soft wearable exoskeleton with compliance tendon-sheath actuation (CTSA) for multi-mode elbow rehabilitation training. The CTSA system is developed on the base of the Hill-based muscle model to imitate the working characteristics of human muscle and improve the compliance and coordination of human-robot cooperation. The FSMCA controller is capable of providing patient-passive and patient-active therapies under different training intensities and encouraging the active participation of the patients with various weakness levels. Further experimental investigations, including the trajectory tracking experiments with/without admittance regulation, the step response experiments with disturbance, the frequency response experiments, and the patient-active training experiments, are carried out by three healthy volunteers. Experimental results demonstrate the effectiveness of the proposed FSMCA control scheme in achieving high control accuracy and favorable frequency response characteristic. Besides, the training intensity can be qualitatively adjusted via appropriate admittance parameters.