Model based open-loop posture control of a parallel ankle assessment and rehabilitation robot

Ankle injuries are very common in daily life out of musculoskeletal or neurological reasons. Traditional ankle therapy usually requires cooperative and intensive efforts from therapists and patients. Robot-assisted ankle rehabilitation techniques have been actively researched in the past few decades. However, limitations exist such as inappropriate robot design, limited range of motion (ROM) and torque generation capability, or lack of measurement of the interaction between robots and patients. This paper proposes a novel ankle assessment and rehabilitation robot (AARR) that could perform a three-dimensional robotic training with real-time ankle assessment. This study focuses on the control of the robot for ankle rehabilitation rather than assessment. A preliminary test on a healthy subject was conducted using the AARR with a model based open-loop controller. Results show that the robotic training is continuous although the trajectory tracking accuracy is not high. It is concluded that the robotic design has potential for ankle rehabilitation. This model based open-loop controller provides continuous robotic training and is suitable for low precision tracking requirement. It can also serve advanced control schemes as the forward feedback.