Determination of the Optimum Berth Number of the Izmir Seaport by Deterministic and Probabilistic Methods Based on Artificial Neural Networks

Seaports are among the vital requirements of a country's trade. The service capacity of a seaport is important to maintain seaport operations. Determining the number of the berthing places is a major decision for planning a seaport by considering the future port capacity. Queuing theory is used to determine the optimum berth number in a seaport. In this study, optimum number of berths of Izmir seaport in Turkey was determined by implementing the queuing theory. Likewise, the ship traffic was predicted using two forecasting models: Linear Regression Analysis and Artificial Neural Networks. Model results were candidates of input variables for the queueing model for estimating the feasible berth numbers between 2015 and 2023. For 2015, feasible number of berthing places was found as 14 and for 2023 it was found as 16 by the queuing theory.