Study On Agent-Based Production Planning And Scheduling System

The stringent market competition and individualization requirements make today' s enterprises facing drastic challenges. To adapt to new environment, enterprises should make full use of human, finance and material resources, hence to hold limited market shares. Just under this background, this dissertation constructs an agent-based Production Planning and Scheduling system (PPS system) integrated with agent concept in computer field. And has an overall and deep study on the application of agent on mechanical engineering. Based on the theoretic research, some related software is developed. And it constitutes a software prototype of cooperative control system of Production-To-Order when combined with concerned efforts of other researchers in my lab.This dissertation has a deep analysis and research on manufacturing system structure, optimized scheduling algorithm, modeling method and programming technology. An idea that builds the PPS system by utilizing agent-based structure and mode is put forward. Along with this idea, an agent-based function model of manufacturing system is formed, and then the agents in the PPS system are grouped. To overcome cooperation difficulties brought by grouping, a method named bridge-communication is introduced. Meanwhile, the physical structure, syntax structure and the KQML communication language of agent and contract net protocol (CNP) are deeply researched. Detailed communicating processes of agents in the system are designed. By doing these works, the concept of multi-agent system based on CNP and KQML is more perfectly.According to the research on product order evaluation and management, production planning and scheduling, a method using flexible process to increase system agility is produced. And related algorithms and simulation experiments are finished. Method proved to be feasible and system runs well.