Studies on modelling and designing of auxiliary robots used in clinical rehabilitation therapy have been done in literature. In these studies, dynamic structures, modelling and kinematic equations of the robot were featured and the control of the systems were realized by using traditional controllers. In this thesis, a parallel mechanism auxiliary ankle rehabilitation robot was designed and control strategies for different rehabilitation exercises were proposed. Different type optimized controllers were designed to minimize the disturbing human effect in trajectory tracking. In addition, a fuzzy logic based adaptive admittance control scheme was proposed to adapt resistance/assistance level provided by the robot according to the patients’disability level. Optimized fuzzy logic, fractional order PID and PID controllers were designed using both particle swarm optimization and cuckoo search algorithms. Additionally, repetitive controller which is suitable for periodic reference trajectories were designed and the performances of all controllers were compared using error-based performance indexes under disturbing effect of robot-human interaction. Furthermore, the effectiveness of the proposed fuzzy logic based admittance control scheme was observed in the experimental results, which were carried out using both adaptive admittance control scheme and static admittance control scheme.
Key Words: Ankle rehabilitation robot, trajectory control, fractional order PID controller, fuzzy logic controller, repetitive controller, adaptive admittance control, PID controller, particle swarm optimization, cuckoo search algorithm, performance indices.