<![CDATA[The existing research about multi-agent systems (MAS) in container terminal have not yet fully depicted the entire operations processes. These past works tend to use static (linear programming-based) modeling methods to model the systems. This method has a weakness because decision optimality is only guaranteed for the agents that are directly involved in the discussion. While the other agents whose performance are affected by the decision, but not directly taking part in discussion, are not considered in the decision-making process. The utilization of MAS-based dynamics negotiation method is also expected to cope lack of the sequential communication methods. Eight agents that depict entities in the terminal are ship, port captain, terminal manager, stevedore, quay crane, straddle carrier, customs, and truck. The agents interact in the processes of ship arrival sequencing, determination of ship’s service time and container picking. Based on the simulation result, agents’ behavior is able to depict the real system. The agents tend to maximize profit in the early period and gradually decrease it to attain common agreement. The velocity of utility function attainment is about 2-5%, but because of the descend monotonic function, negotiation agreement is guaranteed to be attained. Moreover, we provide the detail MAS-based dynamical model as well as the proof that the consensus will be attained by the entire agents.]]>
Copyrights © 2021
