Βαθμονόμηση κυκλοφοριακών μοντέλων προσομοίωσης με τη χρήση του αλγορίθμου SPSA

Μάρκου, Ιουλία; Markou, Ioulia

dc.contributor.author	Μάρκου, Ιουλία	en
dc.contributor.author	Markou, Ioulia	el
dc.date.accessioned	2015-04-28T09:13:34Z
dc.date.available	2015-04-28T09:13:34Z
dc.date.issued	2015-04-28
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/40629
dc.identifier.uri	http://dx.doi.org/10.26240/heal.ntua.4054
dc.description	Εθνικό Μετσόβιο Πολυτεχνείο--Μεταπτυχιακή Εργασία. Διεπιστημονικό-Διατμηματικό Πρόγραμμα Μεταπτυχιακών Σπουδών (Δ.Π.Μ.Σ.) “Γεωπληροφορική”	el
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	Προσομοίωση	el
dc.subject	Βαθμονόμηση	el
dc.subject	Μοντέλο	el
dc.subject	Αλγόριθμος	el
dc.subject	Simulation	en
dc.subject	SPSA	el
dc.subject	Calibration	en
dc.subject	Model	en
dc.subject	Traffic	en
dc.title	Βαθμονόμηση κυκλοφοριακών μοντέλων προσομοίωσης με τη χρήση του αλγορίθμου SPSA	el
dc.title	Calibration of Traffic Simulation Models using SPSA	en
heal.type	masterThesis
heal.classification	Κυκλοφοριακά μοντέλα προσομοίωσης	el
heal.classification	Traffic Simulation Models	en
heal.language	en
heal.access	free
heal.recordProvider	ntua	el
heal.publicationDate	2014-11
heal.abstract	Traﬃc simulation models, with the default parameter values, are not suitable for all studies. The characteristics of the network, the drivers behavior etc. are constantly changing. Each model represents this diversity through its parameters. When the basic format of the model is known, and there is enough data available, calibration is the procedure that aims at estimating the parameters that will lead to results, as close as possible to the observed ones in the ﬁeld. The calibration is a crucial step for the successful application of traﬃc simulation models in transportation planning and real traﬃc management. In this thesis, the calibration methodology using the algorithm Simultaneous Perturbation Stochastic Approximation (SPSA) is studied. The behavior of the algorithm is analyzed through various applications. Several particular behaviors are identiﬁed for which solutions are suggested. The algorithm was prepared to apply even in a quite demanding calibration workload, that of the expressway network of Singapore using the DymaMIT. Results indicate that SPSA successfully manages to reach the optimal solution, however with diﬀerent speed and eﬃciency each time. The parameter values were proved to be very important, especially the parameters a, alpha and c. Mismatches could even lead to the ﬁnal failure of the algorithm’s convergence to the optimal solution.	en
heal.advisorName	Αντωνίου, Κωνσταντίνος	el
heal.committeeMemberName	Σπυροπούλου, Ιωάννα	el
heal.committeeMemberName	Κεπαπτσόγλου, Κωνσταντίνος	el
heal.academicPublisher	Εθνικό Μετσόβιο Πολυτεχνείο. Σχολή Αγρονόμων και Τοπογράφων Μηχανικών	el
heal.academicPublisherID	ntua
heal.numberOfPages	134 σ.	el
heal.fullTextAvailability	true