Dynamic analysis and control system design for a lower extremity exoskeleton

Exoskeleton is a machine composed of a wearable anthropomorphic structure which noticeably magnifies user's might via its actuators. In this research, dynamic modeling and control system design for a lower limb type of this robot were done. In the literature at most a 1 DOF part of the robot is modeled and controlled which doesnt give a good insight on how all of the robot parts are controlled simultaneously. First, a suitable structure was chosen similar to that of UC Berkeley's BLEEX project. Then dynamic equations were derived in sagittal plane using the Newton-Euler method. By an experiment using Xsens system, gate kinematics data were measured and the inverse dynamics was simulated both in SimMechanics and on the model in MATLAB that proved accuracy of the derived model. Impedance control was investigated and some corrective remarks were included in that algorithm. Using this method the robot was controlled. It stabilized the system and the robot followed user's movement exactly. While a load of 50 kilograms was carried, mostly moments of less than 1 (Nm) were applied at each interface among man and robot.