Aristotle University Of Thessaloniki (AUTH)
Ευσταθής λαβή και μεταβίβαση φορτίου αντικειμένου από ρομπότ σε άνθρωπο
Abstract
dc:descriptionThe thesis deals with the problem of grasping by robotic hands as well as the human-robot haptic interaction during object load transfer. A passivity-based stable grasping control law for unknown object weight is exploited and extended in order to be used in the design of two strategies for the safe and natural human-robot object load transfer which takes place in the direction of the gravity field. The controller is extended with respect to the desired grasping force and the object's mass that is allocated to the respective hand. The extension is based on human studies on the grasping and load forces that are developed by the giver and receiver. The first object load transfer strategy is initiated by the giver who linearly decreases its load and grip forces until the full release of the object’s load. It is assumed that the receiver estimates the load transfer successfully and adapts its grip forces accordingly to achieve an efficient object transfer, ie. the receiver is either a cooperative and healthy human or a robot. In case the receiver is not a fully responsive participant and does not or cannot instantly accept the released load, the force exchange is not linear and as a result the object slips from the grasp since the receiver's behaviour is not taken into account. The second strategy is receiver initiated and it involves a rich haptic interaction between the hands. In this strategy, the giver is a stably grasping robotic hand that follows closely the receiver’s lead and ensures haptically that the receiver has stably grasped the object before opening its grip. Therefore, the receiver can be anyone from a fully cooperative robot to an insufficiently responsive human. Theoretical design and validation via simulations are presented for both strategies, demonstrating the advantages of the receiver initiated strategy. This strategy is generalized for cases where the robot's fingers do not possess rolling capabilities on the object's surface and for cases where the load transfer does not take place in the direction of the gravity field. The strategy is also integrated with an arm motion controller to complete the hand-over procedure and is validated by both simulations and experimental results. Consequently, the control law used in the object load transfer strategies is extended for a weightless object in order to achieve control objectives regarding the optimal grasping force which may vary for different kinds of objects or the robot's subsequent desired task following the stable object grasp. This is achieved by controlling the fingers' desired relative orientation in the rolling direction by the a priori optimization of the internal force manipulability ellipsoid. Theoretical design and validation via both simulations and experimental results are presented for the proposed controller, demonstrating its effectiveness. Last, the grasping controller is extended to the three dimensional space where the fingers' desired relative orientation is controlled with respect to two rolling directions on the contact surface. The definition of these directions determines the in-hand object manipulability. Simulation results are presented that validate the theoretical analysis.
Degree
thesis:*- Grantor dc:publisher
- Aristotle University Of Thessaloniki (AUTH)
- Year dc:date
- 2017
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Psomopoulou, Efi
- Ψωμοπούλου, Έφη
Subjects
dc:subject × 12- Ευσταθής λαβή
- Παράδοση αντικειμένου
- Φυσική αλληλεπίδραση ανθρώπου ρομπότ
- Ρομποτικά δάχτυλα
- Stable grasping
- Object hand-over
- Physical human-robot interaction
- Robotic fingers
- Επιστήμες Μηχανικού και Τεχνολογία
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Engineering and Technology
- Electrical Engineering, Electronic Engineering, Information Engineering
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/42252