Research And Practice On Mechanism Of Task Management Driven By Orders In Virtual Enterprise

As the main management pattern for product customization, Virtual Enterprise (VE) realizes resources superiority supplementary in interrelated manufacturing enterprises by the integration of their core abilities. Then the purposes to reduce the product production cost, improve the product quality, shorten the product development time and response to market rapidly can be achieved. Task management driven by orders is the key link to realize the VE management pattern, and includes the trust evaluation of supplier, order task classification, order task allocation and tracking. In this dissertation VE engaged in production customization for mechanical parts is taken as research background, the main contributions and works are as follows:(1) Organization forms of VE and operation modes of outsourcing were analyzed, the management process of orders in VE was researched and the system architecture of task management driven by orders in VE was designed.(2) The trust evaluation method for suppliers was put forward, and includes the multi-expert group decision-making based initial trust evaluation and the trust evaluation based on trading conduct of suppliers. The appraisal steps were given and the method was verified by application examples and simulation tests.(3) Based on Rough Set (RS) attribute reduction, the method of attribute characteristic extraction from order tasks was researched, the order classification algorithm based on improved Self-Organizing Feature Map (SOFM) network was designed and the classification of orders is realized automatically, and they were verified effective by practical specimens. For order task allocation, the bidding strategy of order allocation actively was researched and its corresponding algorithm was designed. On the basis of contract net protocol, the model of order allocation passively was established, and by means of improved Particle Swam Optimization (PSO) algorithm, the method to solve this model was put forward and its performance was analyzed.(4) On the basis of Software as a Service (SaaS) mode and web service, the solution for order tracking across enterprises boundaries was put forward. The data storage and database security in SaaS mode were researched, and the encapsulation of application systems existed in the enterprise for web service was discussed. The order tracking process was designed and the order tracking service was realized based on the Web.(5) The prototype system for product customization of mechanical parts was researched and developed. The classification, allocation and tracking of order tasks in the core enterprise of VE were discussed, and the order management by customers in the system was analyzed. The methods proposed by the author were verified and improved by means of this prototype system in the dissertation.The methods proposed by the author to task management driven by orders in virtual enterprise will have the reference value for similar applications of other manufacturing enterprises.