Summary: | 碩士 === 東海大學 === 工業工程學系 === 89 === Along the quick varying of global markets, many businesses are facing more challenges. The speed of product delivery becomes the key factor to competition for businesses. Supply chain management is an important concept to integrate the resources of both upstream and downstream of a supply chain and facilitate quick and correct response.
Additionally, as global competition rises, the complexity of the problems with which businesses are faced also increases. Traditional individual intelligent agent is not enough to solve the problems. The development of multi-agents then emerges.
In this research, we focus on a central manufactory and treat the following issues: how to use its supply net effectively and integrate its upstream and downstream resources based on adequate information sharing and effective function of order processing.
This paper provides a combined method of Petri-net simulated operation and evaluation and searching function to reach the above-mentioned goals. Four important components are included. First, Petri-net is used to model the supply net of central manufactory and the infrastructure for customer real-time response system. Second, a genetic algorithm is utilized for the optimum search for the results of dynamic simulation resulting at the previous stage. Third, gathering customer preference information provides the foundation of solution search and the suggestions in multi-directions, which fit the customers’ requirements. Fourth, the solution resulting from the optimum operation of GA provides the foundation of product planning and scheduling.
|