Emerging applications will demand robots to deal with a complex environment, which lacks the structure and predictability of the industrial workspace. Complex scenarios will require robot complexity to increase as well, as compared to classical topologies such as fixed-base manipulators, wheeled mobile platforms, tracked vehicles, and their combinations. Legged robots, such as humanoids and quadrupeds, promise to provide platforms which are flexible enough to handle real world scenarios; however, the improved flexibility comes at the cost of way higher control complexity. As a trade-off, hybrid wheeled-legged robots have been proposed, resulting in the mitigation of control complexity whenever the ground surface is suitable for driving. Following this idea, a new hybrid robot called Centauro has been developed inside the Humanoid and Human Centered Mechatronics lab at Istituto Italiano di Tecnologia (IIT). Centauro is a wheeled-legged quadruped with a humanoid bi-manual upper-body. Differently from other platform of similar concept, Centauro employs customized actuation units, which provide high torque outputs, moderately fast motions, and the possibility to control the exerted torque. Moreover, with more than forty motors moving its limbs, Centauro is a very redundant platform, with the potential to execute many different tasks at the same time. This thesis deals with the design and development of a software architecture, and a control system, tailored to such a robot; both wheeled and legged locomotion strategies have been studied, as well as prioritized, whole-body and interaction controllers exploiting the robot torque control capabilities, and capable to handle the system redundancy. A novel software architecture, made of (i) a real-time robotic middleware, and (ii) a framework for online, prioritized Cartesian controller, forms the basis of the entire work.

Motion Control of the Hybrid Wheeled-Legged Quadruped Robot Centauro

LAURENZI, ARTURO
2020-02-13

Abstract

Emerging applications will demand robots to deal with a complex environment, which lacks the structure and predictability of the industrial workspace. Complex scenarios will require robot complexity to increase as well, as compared to classical topologies such as fixed-base manipulators, wheeled mobile platforms, tracked vehicles, and their combinations. Legged robots, such as humanoids and quadrupeds, promise to provide platforms which are flexible enough to handle real world scenarios; however, the improved flexibility comes at the cost of way higher control complexity. As a trade-off, hybrid wheeled-legged robots have been proposed, resulting in the mitigation of control complexity whenever the ground surface is suitable for driving. Following this idea, a new hybrid robot called Centauro has been developed inside the Humanoid and Human Centered Mechatronics lab at Istituto Italiano di Tecnologia (IIT). Centauro is a wheeled-legged quadruped with a humanoid bi-manual upper-body. Differently from other platform of similar concept, Centauro employs customized actuation units, which provide high torque outputs, moderately fast motions, and the possibility to control the exerted torque. Moreover, with more than forty motors moving its limbs, Centauro is a very redundant platform, with the potential to execute many different tasks at the same time. This thesis deals with the design and development of a software architecture, and a control system, tailored to such a robot; both wheeled and legged locomotion strategies have been studied, as well as prioritized, whole-body and interaction controllers exploiting the robot torque control capabilities, and capable to handle the system redundancy. A novel software architecture, made of (i) a real-time robotic middleware, and (ii) a framework for online, prioritized Cartesian controller, forms the basis of the entire work.
13-feb-2020
software architecture; real-time; locomotion; impedance control; wheeled robots; redundancy
File in questo prodotto:
File Dimensione Formato  
phdunige_430776.pdf

Open Access dal 01/06/2020

Tipologia: Tesi di dottorato
Dimensione 21.39 MB
Formato Adobe PDF
21.39 MB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11567/996233
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact