Modeling And Simulation Research Of Coal Enterprise Supply Chain Based On Norm And Multi-Agent

In recent years, some problems have appeared in the coal supply chain such as continuous increasing of inventory, unbalance supply and demand, and the tin ear for the customer requirements. Faced with the dynamic and opening market environment, to understand and meet the needs of the consumers quickly, and adapt to the market environment, customers and suppliers, distributors, retailers need to close the ties of each other, which can dispatch the coal resources on the whole supply chain, to complete favorable negotiations between enterprises. The whole system has formed a complex adaptive system as the change and development of Agent behavior in the Coal supply chain, which affects the evolutional development of the whole coal supply chain. At present an effective way to study the complex adaptive system is modeling and imitation for it combined with multiple Agent theory and technology. Guided by this theory, enterprises and production units inside of the enterprise which are in side of the coal supply chain can be seen as dynamic and variational Agents, internal negotiations between enterprises or among each production unit are actually dynamic negotiations between the intelligent Agents.The Interaction mechanism between Agents can be implemented by dynamic selections of the behavior and resource of the Agent. Contract net which can realize the reasonable allocating for resources and the tasks negotiation between multi-agents is a kind of classic coordination strategy under the complex task. The semantic description tool is dependent to constraint and expression the Agent behavior inside multi-agent system. Norm in the semiotics which has expanded the scope of semantic expression of rules and convention can well implement the constraint and specify of the Agent behavior in the system and well describe the subjective dynamic characteristics of the Agent. Norm as a kind of knowledge the enterprise Agent mastered needs to adapt to dynamic and complex outside environment. GALCS which can realize intelligent learning and automatically updating the rules of the whole organization is a kind of parallel, rule-based intelligent system.This work is derived from the Natural Science Foundation of China (No.70961007), which mainly analyzes the enterprise evolution model and simulation based on norm and multiple-Agent technology. In this paper, negotiation mechanism of Agent is improved with the theory of Norm and multiple Agents on the basis of the analysis of research status on related areas at home and abroad, the main contents are as follows:(1) A kind of enterprise deciding framework is designed based on Norm combined with machine learning classifier system based on genetic algorithm, and an example is given to prove the effective learning and supervision for the enterprise Norm of enterprise model.(2)A global negotiation model is put forward combined with improved genetic algorithm and improved contract net based on Norm and "trust evaluation". The model which uses Norm to describe enterprise information can overall solve the negotiation issues supply chain involved. (3)The previous studies are used in the last simulation experiments. Through the experiment, the effectiveness of the interaction and negotiation mechanism between enterprises Agent is proved.