A multi agent system for the online container stacking in seaport terminals

Abstract

With the continuous development of seaports, problems related to the storage of containers in terminals have emerged. Such problems, referred to in the scientific literature as Container Stacking Problems (CSP) consist in determining the exact location of containers in the storage area of a terminal. Several research works have been conducted to develop systems for the management of container storage operations, which are referred to as Container Terminal Operating Systems (CTOS). Unfortunately, existing systems suffer limitations related to distributed control, online stacking strategies efficiency and their ability to handle dangerous containers. In this paper, we suggest a multi-agent approach for the reactive and decentralized control of container stacking in an uncertain and disturbed environment. A Belief-Desire-Intention (BDI) model has been proposed for the development of the different agents constituting the system. A set of knowledge models and learning mechanisms for disturbance and reactive decision making management are suggested and integrated in the system. The suggested system is able to capture, store and reuse knowledge in order to detect disturbances (as those related to resources breakdown), select the most appropriate storage strategy and determine the most suitable container location.