摘要： 针对连锁零售企业库存量高、库存成本高、配送断货和配送滞后等问题，建立基于多智能体系统的多级库存智能管理系统，应用NETLOGO平台实现多智能体库存优化系统的仿真研究。实验结果表明，该系统可以解决多级库存优化算法不符合实际情况或无法得到结果等问题，可应用于连锁零售多级库存优化和多智能体的建模仿真研究中。

关键词: 人工智能, 连锁零售, 多级库存, 多智能体系统, 建模仿真环境

Abstract: For the severe problems of high inventories, high cost of distribution, goods-stopping, disorderly delivery and distribution lag in retail chain, a multistage inventory control system based on Multi Agent System(MAS) is established and simulated on NETLOGO. It solves two widespread problems in retail chain multistage inventory. The unrealistic optimization algorithms because of hypothesis, and the unsolvable formula because of overmuch variables. In the area of optimization of multistage inventory and research on modeling and simulation of MAS, this system can reach consensus in one step.

Key words: artificial intelligence, retail chain, multistage inventory, Multi Agent System(MAS), modeling and simulation environment

中图分类号: 

• TP18