Iterative Learning Control Strategies for Robot-aided Ankle Rehabilitation

The human ankle plays an important role in daily activities, but it and its actions are vulnerable to physical and neurological injuries. Physiotherapy is labour-intensive and time-consuming, in consequence, the use of robotics to aid ankle rehabilitation has attracted increasing attention. However, currently, robotic equipment is merely an adjunct to the therapist and its automation capability is underutilized. The main technical barrier is that existing control strategies for ankle rehabilitation robots are lack learnability and adaptability while requiring continuous supervision and guidance from the therapist. By the repetitive nature of the rehabilitation scenario, this thesis aims to propose training strategies based on iterative learning control (ILC) to improve the effectiveness of the use of robots in ankle rehabilitation.