Toward a Knowledge Based Multi-agent Architecture for the Reactive Container Stacking in Seaport Terminals

Abstract

In container seaport terminals, one of the most important problems is the one related to the storage of containers. Seaport authorities have invested in means and decision support systems to solve such problems, referred to in this paper as the Container Storage Problem (CSP). Moreover, many unexpected events may occur during the container storage process and, consequently, scheduled position of containers must be modified. Although the number of developed Decision Support System (DSS) for the management of container storage, there is still a need for DSS able to deal different type of disturbance simultaneously. In this paper, we suggest a set of knowledge based DSS for the distributed control of container storage process in an uncertain and disturbed environment. The suggested system is based on a set of knowledge models and learning mechanisms which are integrated in a multi-agent system. Numerical experiments show that knowledge based systems combined with Multi-Agent Systems seems be effective for the real time Container Storage in seaport terminals.