Research On Real-time Container Terminal Scheduling System Based On Agent

As wireless communication, mobile terminal, distributed data processing information technology in container terminal is becoming more and more popular, a new kind of Container Terminal Scheduling model of Real-time is put forward. Compared with the traditional scheduling model it is more useful to improve efficiency of resource utilization and terminal operation. The researches on Container terminal resources optimization and dispatching management involve management, operations research, information science, port engineering and many related fields. So Container terminal real-time scheduling system is the integration of theory and application of technology. The study in this thesis can play an active role in the application, perfection, popularization of the theory of container terminal operation system, the theory of combinatorial optimization, application framework of Mulit-Agent and Mobile-Agent. Based on Agent Oriented Analysis using AUML, the function and interactive model of container terminal real-time scheduling system were designed in this thesis. Applying this model to further improve the utilization of productive resources, to strengthen cooperation between operations, to improve the working efficiency and throughput capacity of wharf, to guarantee the stability and continuity of terminal operations, to improve customer satisfaction has practical significance.Mainly studies completed in the paper include:(1) Researched on real-time scheduling framework model based on AgentBased on the method of AUML AOA, the Role-Active model, the Structural Framework Model, and the interactive model were designed. These models analyzed the functions, the relation between functions and the interactive mode in Real-time scheduling system from the macroscopic Angle, and all these models macroscopic Angle made up system framework model.(2) Designed Real-time scheduling oriented intelligent Plan AgentsReal-time oriented, supporting dynamic adjustment, intelligent operation planning methods were studied. In this thesis Berth-crane plan Agent model, Ship plan Agent model, Machine plan Agent model were designed in detail. And relevant optimization algorithms were put forward. The thesis also summarized how Agents collaborative work to realize real-time scheduling for emergency, and optimization methods of acquiring Agent environmental data.(3) Designed Real-time operation model by cooperation with shore crane, truck based on Mobile-Agent technologyMobile-Agent technology has many advantages such as save network bandwidth, overcome the network latency, support asynchronous independent execution, with dynamic adaptability, enhance application robustness and tolerance, etc. These advantages are very important for real-time scheduling in wireless-network environment. The thesis provided a real-time operation model based on mobile-agent and give the detail explanation of architecture and working mechanism, discuss the key technology in design and implementation. This model can reduce influence of the quality of wireless network communication on system, improve the system response ability and offer a new method to implement real-time dispatcher in container terminal.Generally speaking,this thesis provided the design of real-time scheduling model, and gave the examples of supporting "unloading with loading" and "heavy come heavy go" operation mode in real-time scheduling system. It can provide a comprehensive reference model for future development of container terminal.