Prethodno priopćenje
https://doi.org/10.7307/ptt.v27i5.1689
Setting the Port Planning Parameters In Container Terminals through Bayesian Networks
Tomás Rodríguez García
; Polytechnic University of Madrid, SPAIN
Nicoletta González Cancelas
; Polytechnic University of Madrid, SPAIN
Francisco Soler-Flores
; Polytechnic University of Madrid, SPAIN
Sažetak
The correct prediction in the transport logistics has vital importance in the adequate means and resource planning and in their optimisation. Up to this date, port planning studies were based mainly on empirical, analytical or simulation models. This paper deals with the possible use of Bayesian networks in port planning. The methodology indicates the work scenario and how the network was built. The network was afterwards used in container terminals planning, with the support provided by the tools of the Elvira code. The main variables were defined and virtual scenarios inferences were realised in order to carry out the analysis of the container terminals scenarios through probabilistic graphical models. Having performed the data analysis on the different terminals and on the considered variables (berth, area, TEU, crane number), the results show the possible relationships between them. Finally, the conclusions show the obtained values on each considered scenario.
Ključne riječi
containerised traffic (trade); Bayesian networks; planning; forecast; port capacity
Hrčak ID:
149614
URI
Datum izdavanja:
28.10.2015.
Posjeta: 1.355 *