Improvement of Performance of Actuator with Adjustable Stiffness (AwAS) by Changing the Structure of the Mechanism and Using Sliding Mode Control Methodology

EasyChair Preprint 8286

Abstract

In recent years, due to the difficulty of rehabilitation exercises, rehabilitation robots have been used to assist physiotherapists, and due to the many interactions between the patient and the robot, robot safety is one of the most critical issues related to rehabilitation robots. In this paper, the mechanism of a variable stiffness actuator called AwAS is improved. Variable stiffness actuators are used in robots that have a lot of physical interactions with humans. The AwAS actuator has two motors. The first one controls the position and the second one controls the stiffness. The stiffness of the AwAS actuator is changed by changing the position of the springs embedded in it. The AwAS mechanism is such that the force required to move the springs to change the stiffness is approximately perpendicular to the force produced by the springs, ie the direction of their change in length. If these two forces are precisely perpendicular, the force required to move the springs will be minimized. In this paper, by changing the mechanism, the two forces are always perpendicular and less force is needed to move the spring. In this system, a sliding mode control methodology is used to prevent the robot vibrations. The robot has three degrees of freedom, and the dynamics of the system indicate that the system is nonlinear and has three inputs and three outputs that affect each other; that is, they are coupled together. Therefore, a method is used that is used for MIMO systems.

Keyphrases: AWAS, MIMO, control, robot, sliding mode, variable stiffness actuator

@booklet{EasyChair:8286,
  author    = {Hassan Mohammadi and Habib Ahmadi and Hamid Aris},
  title     = {Improvement of Performance of Actuator with Adjustable Stiffness (AwAS) by Changing the Structure of the Mechanism and Using Sliding Mode Control Methodology},
  howpublished = {EasyChair Preprint 8286},
  year      = {EasyChair, 2022}}