4th International Conference on Intelligent and Fuzzy Systems (INFUS), Bornova, Turkey, 19 - 21 July 2022, vol.504, pp.352-359
Maritime authorities not only have ensured reactive but also proactive measures in the straits, canals and narrow waterways with geographical restrictions and high traffic density to mitigate possible accident risks. These measures include a variety of approaches to conducting a realistic risk analysis. Considering the previous accidents, increasing traffic density and unique difficulties in the Turkish Straits, taking proactive measures to ensure their safety has become an important issue. Furthermore, a wide variety of ships navigating this waterway have been having difficulties in making good judgments. Therefore, this study has focused on the clustering of the ships in the Turkish Straits which can be deemed as the first step of realistic risk analysis in narrow waterways. Fuzzy C-Means clustering method has been employed, based on Sailing Plan-1 reports data between 2005 and 2021 in order to reveal maritime traffic characteristics for further analysis. Results have shown that three clusters are suitable for ship risk profile as a first step but an additional hierarchical layer may be needed to overcome the contradictory situations.