Development of a human-robot dynamic model to support model-based control design of an upper limb rehabilitation robot

There are a large number of people with movement disorders who have difficulties doing their daily tasks independently. Much research has been devoted to the therapy of these people, and rehabilitation robots have been developed to assist with therapy. In robotic rehabilitation, since the human body is interacting with a mechanical device, safety issues in the design of appropriate control strategies are very important. Thus, rehabilitation robots usually use control approaches that consider the human body interacting with robot as a mechanical impedance [1]. Characteristics of this mechanical impedance may vary depending on different musculoskeletal factors such as posture and muscle contraction dynamics [2], and multiple experiments are required to evaluate them. However, in the current state of the art, rehabilitation robot controllers conservatively assume the robot interacting with some general static impedance models. In other words, there is a lack of research that considers realistic human body interactions with the rehabilitation robot. Since this interaction affects therapy procedures, the objective of the current study is to develop an integrated human-robot dynamic model with real-time simulation capability to support model-based robot controller design. Upper limb motor defects are common among stroke patients, so this study is focused on an upper extremity rehabilitation system.