A three-DoF upper limb exoskeleton's design, modeling, and interaction-based control

Abstract

For the last ten years, human-coordinated robots are among the innovative technological designs that facilitate human life, especially in the defense industry and health fields, and attract the attention of researchers. This article focuses on the three-dimensional design of a three-degrees-of-freedom (DoF) ridgeback upper extremity exoskeleton robot, mathematical modeling of the system, IMU-based human-robot interaction architecture and joint controller design. The kinematic model of the 3 DoFs upper extremity exoskeleton arm was obtained by the Lagrange method. Then, the multi-body system model of the robot arm was created and its dynamic behavior was analyzed with motion simulations. A fractional PID was designed to control the joint angles of the exoskeleton robot and the coefficients of the controller were determined by optimization. The performance of the proposed system and the controller has been tested under different load conditions of 1.825, 3.650, 6.860 kg both in the simulation model and in the test system manufactured and presented in comparison. Although the maximum tracking error of the fractional PID controller increases proportionally with the amount of weight carried, the average tracking error for each joint 2.2401 & DEG;, 3.252 & DEG;, and 4.471 & DEG; under maximum load (6.860 kg), the instant maximum tracking error for each joint 8.496 & DEG;, 8.855 & DEG;, and 10.8666 & DEG; the mean tracking error time is 0.182 s, 0.278 s, and 0.211 s, respectively. Although the performance of the real test system is relatively poor, it is seen that the tracking error with the proposed IMU interaction fractional controller system is very low compared to the upper extremity robots in the literature. Graphics and mathematical results show that the proposed method is applicable to the defense industry and rehabilitation robots.