基于多Agent的生产线建模、优化与仿真研究

发布时间：2018-07-07 15:42

本文选题：系统建模与仿真 + 多Agent技术　；参考：《合肥工业大学》2017年硕士论文

【摘要】：随着互联网技术、信息技术、人工智能技术以及现代管理技术地不断进步,现在制造系统开始往网络化、信息化与智能化的方向发展。经济市场的快速变化,消费者多变的需求,使先进制造系统也变得更加复杂、多变。从经济性、安全性等角度考虑,依靠计算机仿真技术描述真实生产系统,在虚拟环境下验证生产线优化决策,提高生产线的生产效率,是一种可行的手段。Agent具有智能性、自主性、交互性,基于多Agent的建模与仿真方法是一种新的建模与仿真方法,能够将复杂的制造系统分解简化为一个个独立的子系统来解决,这种建模方法能够很好描述系统的微观行为以及涌现的宏观现象。本文以玻璃堆垛系统为背景,利用多Agent技术进行模型设计,基于有限状态机原理建立仿真模型,通过计算机仿真技术与P2P网络通信技术相结合,实现了玻璃堆垛的多Agent系统的分布式仿真过程,并且经过仿真分析,观察与统计仿真数据,提出生产线优化策略,为决策者提供理论支持。论文首先研究了多Agent的理论知识,包括Agent的基本结构、多Agent的交互与通信机制以及多Agent建模的方法与步骤。然后分析了玻璃堆垛系统的真实模型以及相关数据的采集,通过Simulink/Stateflow仿真工具搭建模型并进行实验研究,根据实际生产情况,提出相关控制策略。VC编写通信程序,完成了多Agent系的分布式仿真。同时通过仿真数据的统计与分析,验证生产线优化策略。
[Abstract]:With the continuous progress of Internet technology, information technology, artificial intelligence technology and modern management technology, manufacturing system is developing towards the direction of network, information and intelligence. The rapid changes in economic markets and the changing needs of consumers make advanced manufacturing systems more complex and changeable. Considering from the point of view of economy and security, it is a feasible means to describe real production system by computer simulation technology, to verify the optimal decision of production line in virtual environment and to improve the production efficiency of production line. Autonomous, interactive, multi-Agent based modeling and simulation method is a new modeling and simulation method, which can decompose complex manufacturing system into independent subsystems to solve the problem. This modeling method can well describe the micro-behavior and emerging macro-phenomena of the system. In this paper, the glass stacking system is used as the background, the model is designed by using multiple Agent technology, and the simulation model is established based on the theory of finite state machine. The computer simulation technology is combined with the P2P network communication technology. The distributed simulation process of glass stacking multi-Agent system is realized, and through simulation analysis, observation and statistical simulation data, the optimization strategy of production line is put forward, which provides theoretical support for decision makers. This paper first studies the theoretical knowledge of multiple Agent, including the basic structure of Agent, the interaction and communication mechanism of multiple Agent, and the methods and steps of multi-Agent modeling. Then the real model of glass stacking system and the collection of relevant data are analyzed. The model is built by Simulink / Stateflow simulation tool and the experimental study is carried out. According to the actual production situation, the relevant control strategy is put forward. VC is used to write the communication program. The distributed simulation of multiple Agent systems is completed. At the same time, the optimization strategy of production line is verified by statistics and analysis of simulation data.
【学位授予单位】：合肥工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.9

【参考文献】