Research on Optimization Model and Algorithm of Port Cluster System

Based on multi-agent simulation method in complex system theory, it is proposed in this paper the study route, basic framework and research approach of layered optimization for port cluster, to explore the structure of intelligent agents and its competitive and cooperative relationships. Global optimized intelligent agents are set up to optimize the general layout of port cluster and co-optimize the intelligent agents. Besides, the scale effect factor of port, based on the scale effect is introduced in this paper to build the sharing model of port-hinterland cargo freight volume. The factor is applied in switching rules of port intelligent agents. The design takes the high-efficient algorithm which is based on multi-agent Genetic Algorithm. The feasibility and efficiency of the model has been demonstrated through empirical analysis. The theory, method and results of this research could provide theoretical basis and methodological guidance for optimization of port cluster system.