logo SBA

ETD

Digital archive of theses discussed at the University of Pisa

 

Thesis etd-05062024-114619


Thesis type
Tesi di laurea magistrale
URN
etd-05062024-114619
Thesis title
Design, implementation and testing of a residual-enhanced adaptive backstepping control method for a variable stiffness humanoid robot
Department
INGEGNERIA DELL'INFORMAZIONE
Course of study
INGEGNERIA ROBOTICA E DELL'AUTOMAZIONE
Supervisors
.
relatore Prof. Bicchi, Antonio
relatore Prof. Grioli, Giorgio
Keywords
  • Adaptive Backstepping Control
  • Alter-Ego
  • Compliant Robotic Joints
  • Soft Robotics
  • Variable Stiffness Actuation
  • VSA-CubeBot.
Graduation session start date
06/06/2024
Availability
Withheld
Release date
06/06/2094
Abstract (Inglese)
Abstract (Italiano)
In recent years, there is a growing interest in the robotics community in robots with elastic joints. One of the main advantages is the ability to interact safely with humans. Due to their softness and ability to absorb shock, these robots reduce the risk of injury during physical interaction with people. This feature makes them particularly suitable for situations where close cooperation between
humans and machines is required, such as in shared work environments. In addition to safety, robots with elastic joints show remarkable adaptability to environmental changes. Their flexibility enables them to handle unexpected situations more effectively than rigid robots. However, the control of variable stiffness robots is challenging due to their highly non-linear behavior. In this thesis, four different dynamic control techniques are developed and applied to the arms of Alter-Ego, a mobile system with a functional anthropomorphic upper body powered by variable stiffness actuators. The controllers were validated and tested in Matlab/Simulink, in the open-source Gazebo simulation environment, and finally in real robot. Next, a campaign of experiments was conducted involving several participants in order to evaluate the actual usability of the new control system and compare it with the one already implemented in order to determine user preferences.
File