Container traffic prediction in a seaport stockyard implementing artificial bee colony algorithm compared with the genetic algorithm

Abstract

Container traffic prediction in a seaport stockyard implementing artificial bee colony algorithm compared with the genetic algorithmAfter finishing the construction stage of a container terminal, it is very difficult to modify the current port configuration and properties. For this instance, to avoid of the sunk costs, it is crucial for the seaport designers to predict the container traffic at any stage of the design phase. One of the important facilities of a seaport is the container stockyard in which the containers are held for handling operations. The preliminary design of this facility directly influenced by the container traffic and the probable container amount at the stock area. Hence, the required size of the stock area is determined and the further construction costs are directly related with this design parameter. This study proposed a forecasting model for obtaining the daily value of the containers in the stockyard. The proposed model is a simple regressed model with the past records of container traffic in the storage area. The model results showed that, artificial bee colony algorithm successfully predicted the model coefficients even better than genetic algorithm. The results indicated that the model performance is sufficient for predicting the contain amount in a seaport considering the observed records.