Βέλτιστη επιλογή σημείων κράτησης για συνεργατικά υποθαλάσσια οχήματα με βραχίονα

Τάραντος, Σπυρίδων; Tarantos, Spyridon

dc.contributor.author	Τάραντος, Σπυρίδων	el
dc.contributor.author	Tarantos, Spyridon	en
dc.date.accessioned	2018-01-22T11:54:24Z
dc.date.available	2018-01-22T11:54:24Z
dc.date.issued	2018-01-22
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/46247
dc.identifier.uri	http://dx.doi.org/10.26240/heal.ntua.14764
dc.rights	Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα	*
dc.rights.uri	http://creativecommons.org/licenses/by-nc-nd/3.0/gr/	*
dc.subject	Robotics	en
dc.subject	UVMS	en
dc.subject	Grasp points planning	en
dc.subject	Cooperative manipulation	en
dc.subject	Επιλογή σημείων κράτησης	el
dc.subject	Υποθαλάσσια οχήματα με βραχίονα	el
dc.title	Βέλτιστη επιλογή σημείων κράτησης για συνεργατικά υποθαλάσσια οχήματα με βραχίονα	el
dc.title	Optimal grasp points selection for cooperative underwater vehicle - manipulator systems	en
heal.type	bachelorThesis
heal.classification	Ρομποτική	el
heal.classification	Robotics	en
heal.classificationURI	http://data.seab.gr/concepts/989f7eb5ce955dbf96b4eebf1ff0aaec33f7e858
heal.classificationURI	http://data.seab.gr/concepts/989f7eb5ce955dbf96b4eebf1ff0aaec33f7e858
heal.language	el
heal.language	en
heal.access	free
heal.recordProvider	ntua	el
heal.publicationDate	2017-10-20
heal.abstract	In recent years, robotics have become an increasing ﬁeld of study with rapid growth. This mainly happens due to their applicability in everyday life. A great characteristic of robotics is their ability to give access to areas where humans would not be able to reach or even if they would, they would not be able to execute long lasting operations, there. An area like this is the ocean, where many long duration operations are taking place, making the use of robots necessary. This thesis focuses in the use of Underwater Vehicle - Manipulator Systems (UVMS), that are usually suitable for operations like these. The application that is studied is a cooperative pick-and-place operation in which a team of UVMSs have to reach an object, grasp it and transfer it from its initial location to a ﬁnal one. More speciﬁcally, this work is dedicated to the selection of the grasp points where the UVMSs’ end-eﬀectors have to grasp at in order to lift, manipulate and transfer an object from its initial location to a ﬁnal one. A main characteristic, of applications like this, is that the environment, where the robots have to execute their tasks, is unstructured and likely unexplored. But even if we know the exact environment’s structure a priori, it is possible to contains moving obstacles with unpredicted motion that might interrupt predeﬁned tasks. For these reasons, in this work, for the evaluation and the selection of the grasp points, non-task speciﬁc grasp quality measures are used. As a result, the quality measures do not aim to an optimal execution of a set of predeﬁne posterior tasks, but to the system’s (UVMSs and manipulated object) ability to execute each task, that might arise during the operation, in the best possible way. In this work, two novel non-task speciﬁc quality measures for the selection of optimal grasp points on an object in a cooperative pick-and-place operation by a team of UVMSs, are presented. These measures are extracted from the analysis of the system’s dynamic manipulability ellipsoid (DME). The system’s DME is used as a mapping from the system’s control input space to the system’s acceleration space connecting the consuming energy by the UVMSs’ actuators with the provoked acceleration on the manipulated object. As a result the two proposed measures aim in the maximization of the system’s ability to accelerate the object by also minimizing the consuming energy for this purpose. Each measure achieves this goal in a diﬀerent way. The ﬁrst proposed measure is the volume of the system’s DME. This measure aims in the maximization of the DME in every direction. Generally, this measure does not take into account the acceleration produced by the system’s weight, for this reason it is proposed to be used combined with a constraint that, based in this acceleration, guarantees a bound in the system’s minimum performance, as concerns the acceleration of the object’s center of gravity. As concerns the second measure, this is the minimum distance in the translational and rotational acceleration space, as arise from the decomposition of the system’s dynamic manipulability ellipsoid.This measure provides grasp points that guarantee the maximum possible minimum system’s performance as concerns the acceleration of the object’s center of gravity. More speciﬁcally, it is guaranteed that the system will be able to accelerate the object in the most diﬃcult translational and rotational direction in the best way, i.e. higher magnitude with lower energy consumption. In order to select the grasp points, the proposed grasp quality measures have vii to be embodied in an optimization scheme, whose objective function will be this very measure. For this reason, two optimization schemes were created for the purpose of this thesis, one for each measure. As constraints are established the limitations imposed be the UVMSs, i.e. joint limits and actuators’ maximum torque, and the shape of the manipulated object. Finally, in order to clarify the proposed measures and to verify their eﬃciency, the optimization schemes were solved for various case studies, i.e. diﬀerent number of UVMSs and object of varied shapes. The results were analyzed in order to understand the advantages and disadvantages of the proposed measures. A comparison between the proposed measures is taking place in order to illustrate their diﬀerences due to the diﬀerent way that each of the two measures maximizes the system’s DME.	en
heal.advisorName	Κυριακόπουλος, Κωνσταντίνος	el
heal.committeeMemberName	Κυριακόπουλος, Κωνσταντίνος	el
heal.committeeMemberName	Παπαδόπουλος, Ευάγγελος	el
heal.committeeMemberName	Πουλακάκης, Ιωάννης	el
heal.academicPublisher	Εθνικό Μετσόβιο Πολυτεχνείο. Σχολή Μηχανολόγων Μηχανικών. Τομέας Μηχανολογικών Κατασκευών και Αυτομάτου Ελέγχου. Εργαστήριο Αυτομάτου Ελέγχου και Ρυθμίσεως Μηχανών και Εγκαταστάσεων	el
heal.academicPublisherID	ntua
heal.numberOfPages	83 σ.	el
heal.fullTextAvailability	true