Implementation Of Supply Chain Simulation System Based On Agent

In a competitive market, each company wants to send the accurate number and right quality of products to customer in right place and at right time with lowest cost. However in the global supply chain management, because each supplier responds to the short life-cycle and quick response need for the terminal products, all suppliers have to reduce the stock and enlarges the demand from customer. A phenomenon called the Bullwhip effect is making the excessive products and shortage of materials. Therefore, how to adopt the valid inventory policies to satisfy stochastic customer demand is the most important thing of the supply chain management.To illustrate and evaluate the issues related to the research, we implement the single supply chain system by simulating a MIT Beer game. First, we construct the supply chain structure and inventory policies of facilities, the optimal inventory policy is found by computer simulation. We develop an intelligent agents system to search optimal inventory policies. In particular, to find policies that mitigates the impact of the Bullwhip effect.In order to reduce the life cycle of supply chain simulation project and reuse the model we developed before, this paper provide a new reusable design pattern which is different from the problem solving strategy. At first, this paper provides the division of Conceptual Model which includes the division of entity Agent and Function Agent. The Function Agent is further divided into different Agents such as Order entry management Agent and Procurement management Agent, ect. On the basis of division of the Conceptual Model, Experimental Model is developed. The reuse of Experimental Model is the vital of accelerating the development of the supply chain simulation system. At the end of the paper, a supply chain system is provided under the environment of eM-Plant. Beer game is also simulated under this system.