Research And Implementation On Agent Based Material Scheduling And Transporting System For Armed Police Forces

In recent years, The People's Armed Police Force (PAPF) logistics department had strengthened information construction greatly. Especially on infrastructure construction, development of information standards, system integration among applications. They had accomplished lots of work and made notable achievements. However, the information systems we are using is gradually exposing some weaknesses. For example, the system does not responds timely to their environment, computers that handling the core business logic process will bring in performance bottlenecks, the feedback information was not processed sufficiently and so on. In fact, modern logistics information system works in a distributed dynamic environment, which is composed by large amount of computing nodes, they interact with each other continuously, the traditional centralized approach had not met this environment's characteristic.Agent technology is an effective method to solve large-scalable distributed problem in dynamic environment. It is called Agent Oriented Software Engineering (AOSE) that using agent technology to build multi-agent system. AOSE is the theory to guide the development of multi-agent System(MAS). We use the Gaia theory to analyze and design the system, and take Cougaar platform as the development tool.We begin with the Requirements Analysis of the PAPF logistics Material Scheduling and Transporting System (MSTS), and gain the general demand and Functional requirements. We studied the Gaia theory and Cougaar platform as well as processing system analysis and design with Gaia theory , abstract the key role of the MSTS, and then obtains the Agent model, Service model and acquaintance model of the system. Contract Net protocol is used to guide the agent collaboration with each other to accomplish works, processing flow of Material Scheduling sub-system in AMSTS is also promoted based on this protocol. Towards the The Shortest Path problem in Material Transporting sub-system, this paper integrates partition-based speed up of Dijkistra Algorithm with Agent, and uses multi RMA to cooperatively find the solution. In the code stage, we address each-to-each relationship between Gaia model and Cougaar components in detail, we promotes a method on structuring the social organization of the system ,develops the service model based on the task processing strategic of cougar, derive new acquaintance model from combing the KQML, which can be used to guide the coding work. Finally, we development a prototype system, test the effectiveness of Gaia-Cougaar integrated software engineering methods in the construction of multi-agent system, furthermore, we find that the new AMSTS performs well..