This thesis aims to propose a multi-agent based hierarchical model for the operations of container terminals. We have divided our model into four key agents that are involved in each sub processes. The proposed agent allocation policies are recommended for different situations that may occur at a container terminal. A software prototype is developed which implements the hierarchical model. This web based application is used in order to simulate the various processes involved in the following operations on the marine side in a case study of a container terminal in Sweden by adopting a multi-agent based simulation technique. Due to the increase in usage of container transportation, container terminals are experiencing difficulties in the management of the operations. The software provides a decision support capability to terminal managers for scheduling and managing the operations effectively while also visually presenting the time it takes to complete the process and its associated cost. Terminal managers need to implement certain policies to improve the management and operations of the container terminal. The policies are evaluated and tested under various cases to provide a more comparative overview. The results of the simulation experiments indicate that the waiting time for arriving vessels is decreasing when in queue with more than three vessels arriving on same day.

Simulation is a good technique that help analyst to take decision considering each factor of problem that is to be simulated. Simulation in addition with multi agent environment provides better understanding of modeling the entities. The complexity of CT environment and multiple involvement of agents simultaneously enables CT suitable domain for multi agents. We have modeled the four CT operations that are carried out at each CT. These operations are modeled in hierarchical sequence like berth allocation, QCs allocation, Transport vehicles allocation and YCs allocation. The most important factor of simulation is the measurement of dynamic time of each operation. We have simulated and compared different agents active time and service time compared with associated cost. The berth allocation is very important asset from all the operations that are carried out at the CTs. The effective utilization of FCFS berth allocation policy reduces the vessel waiting time in waiting queue. The developed terminal simulator tool allocates all resources dynamically while looking at the number of containers that will be loaded and unloaded at QS and yard storage area. The result of simulation tool presents the good dynamic allocation of transport vehicles. The dynamic resource allocation helps to minimize the congestion and bottlenecks that may occur at CTs. The result of three experiments depicts that the berth allocation and agent allocation is improved and vessel service time is reduced at berth side which automatically reduces vessels waiting time in queue. Besides this the transport allocation and YCs allocation is dynamically assigned by looking at number of containers in the vessels. The terminal simulator helps managers to analyze the simulated results and take better decision at hand.