The Research Of Multi Agent-Based Agile Scheduling Of National Economy Mobilization

Mobilization of National Economy is such a series of activities those enhance the speedy response when National Economy meet an emergency and change the economy condition from peaceful time to war time, in order to protect the national security and meet the war demand. It is the agile and logical mobilization which involves the steps of producing, scheduling, scheming and transporting on arm, material and so on. The requirement variety of the war and the emergency results in the widely area and complex work. Agile scheduling is an important process of dynamic alliance to achieve produce management corporately. The scheduling on the National Economy Mobilization has the same characteristic of distributing, multi-objective and dynamic as it on dynamic alliance in peaceful time, but more differences on agile, time-limited and so on. How to model such a distributed emergency scheduling system and to solve the questions of task allocation, member negotiation, is an important researching field. With the artificial intelligence, especially distributed artificial intelligence development, MAS has become a particular method to research scheduling problem. This thesis is mainly from the project of simulation system of Province-level National Economy Mobilization. It will research the produce agile scheduling question in of National Economy Mobilization base on the Multi-Agent theory. Firstly, the thesis introduces the principle theory of agent, MAS and agent modeling. Secondly, it discusses the characteristic of National Economy Mobilization and the necessity and significance of the system. The discussion is focused on the agile scheduling in National Economy Mobilization and the research method. Thirdly, this thesis uses Multi-Agent method to carry on research to this problem. It has proposed produce agile scheduling distributed architecture, and set up Multi-Agent negotiation framework based on the contract net. At the same time it introduces task acquaintance and multi-attribute decision making, which improve the contract net model further. On these foundations, this thesis proposes the most important methods which are the Multi-Agent negotiation model based on improved contact net and its task scheduling optimization strategy. This model has offered very good solution for the fact that the pharmacy agile scheduling of national economy mobilization case. And it can better settle the uncertainty, dynamics and complexity of national economy mobilization and meet the demands of agile scheduling.