DISTRIBUTED PLANNING IN CELLULAR FLEXIBLE MANUFACTURING SYSTEMS.

This paper presents a novel approach concerning the design of distributed information systems in a computerized manufacturing environment, which consists of loosely coupled manufacturing cells. The planning and control in this environment is related to distributed problem solving, an important artificial intelligence problem. Because the control is distributed, the key issue is to organize the interaction and information flows among cells for task sharing and resource allocation. Structured as the negotiation process, the information flows among cells can be regulated by a distributed scheme. This scheme is specified by a combination of the production rules and the Petri net model, implemented in a rule-based system. Issues of implementation and algorithmic efficiency are also discussed.