Μοντελοποίηση ατυχημάτων σύγκρουσης (collision) στις θαλάσσιες μεταφορές με χρήση της μεθόδου Bayesian Networks

Παναγιωτόπουλος, Φαίδων-Αιμίλιος-Νίκος; Panagiotopoulos, Faidon-Aimilios-Nikos

dc.contributor.author	Παναγιωτόπουλος, Φαίδων-Αιμίλιος-Νίκος	el
dc.contributor.author	Panagiotopoulos, Faidon-Aimilios-Nikos	en
dc.date.accessioned	2015-09-18T10:52:19Z
dc.date.available	2015-09-18T10:52:19Z
dc.date.issued	2015-09-18
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/41322
dc.identifier.uri	http://dx.doi.org/10.26240/heal.ntua.10061
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	Μοντελοποίηση ατυχημάτων	el
dc.subject	Accident modeling	en
dc.subject	Δίκτυα πίστης	el
dc.subject	Ατυχήματα σύγκρουσης	el
dc.subject	Θαλάσσιες μεταφορές	el
dc.subject	Ανθρωποκεντρικός σχεδιασμός	el
dc.subject	Accident modelling	en
dc.subject	Human centered design	en
dc.subject	Bayesian networks	en
dc.subject	Collision accidents	en
dc.subject	Maritime transports	en
dc.title	Μοντελοποίηση ατυχημάτων σύγκρουσης (collision) στις θαλάσσιες μεταφορές με χρήση της μεθόδου Bayesian Networks	el
dc.title	Collision accident modelling in maritime transports with use of the Bayesian Networks method	en
heal.type	bachelorThesis
heal.classification	Maritime transports	en
heal.classification	Θαλάσσιες μεταφορές	en
heal.language	el
heal.language	en
heal.access	free
heal.recordProvider	ntua	el
heal.publicationDate	2015-07-24
heal.abstract	Τhis diploma thesis concerns an integrated study of ship collisions based on the influence of the human element and its consisting factors in the accident. Aim of this dissertation is the collision of ships probability estimation and the human factors leading to it. For this reason, a literature review concerning appropriate methods for accident modeling (and other industries) is presented which is the main theoretical part of this thesis. In the English part, we refer to the importance of the human factor in maritime accidents and the need to design human based systems. Firstly, the principles of Human Centered Design for ships are thoroughly examined, in order to elicit the elements that effect human performance. Afterwards the Bayesian Networks method is selected and various versions are created in both a generic and a more detailed level. Then, we begin to build the model in 2 stages: First we make the frame of the model and then we optimize it, adding elements about the human factors from TRACEr method. All the above are described in detail. After the construction of the defining evidence node wherever we want, we see the change of the final probability of collision. Once the construction of the BN, we define the probability tables of all nodes. Five case studies are extracted from database, in order to run the model, extract the results and even improve it. . Each case study is presented, evidences are selected from their narrative and running the BN model based on these evidences conclude in several probability results for performance and collision. A result presentation and a sensitivity analysis follow for all target nodes. In conclusion, factors that have a strong influence on human performance and collision are selected and highlighted for further research for human performance improvement and collision avoidance.	en
heal.advisorName	Βεντίκος, Νικόλαος	el
heal.committeeMemberName	Βεντίκος, Νικόλαος	el
heal.committeeMemberName	Λυρίδης, Δημήτριος	el
heal.committeeMemberName	Γκίνης, Αλέξανδρος-Αλβέρτος	el
heal.academicPublisher	Εθνικό Μετσόβιο Πολυτεχνείο. Σχολή Ναυπηγών Μηχανολόγων Μηχανικών. Τομέας Μελέτης Πλοίου και Θαλάσσιων Μεταφορών	el
heal.academicPublisherID	ntua
heal.numberOfPages	305 σ.
heal.fullTextAvailability	true