A Study on the Estimation of Optimal Traffic Distribution near Breakwater in Busan Port

ABSTRACT Many studies have been carried out regarding the navigation algorithms and scenarios of other and own ship related to autonomous ships. However, there was a lack of research on the selection of safety ranges during passing marine structures. In this study, marine traffic analysis was performed at Busan Port in June 2020. As a result of analyzing optimal distribution group through Cullen and Frey graph, it was selected as gamma distribution, lognormal distribution, and normal distribution. The statistic was analyzed by the ranking of the goodness-of-fit test according to Kolmogorov-Smirnov test, Anderson-Darling test, Cramer-von Mises test, Akaike’s Information Criterion, Bayesian Information Criterion. As a result, it was analyzed that lognormal distribution at arrival and gamma distribution at departure was best fitted. In this study, the optimal distribution was derived based on Bayesian Information Criterion through observations of 32 distributions through Monte Carlo simulation, and the results were the same. In general, when passing through the route and TSS, we performed the study by assuming the distribution as normal distribution. However, it is expected that more accurate traffic distribution can be derived based on this study, since distribution is different depending on the characteristics and circumstances.