China Mechanical Engineering

Previous Articles     Next Articles

Multi-Agent Job Shop Scheduling Strategy Based on Pheromone

CHEN Ming1;ZHU Haihua1;ZHANG Zequn1;JIN Yongqiao2;WANG Yingcong1;TANG Dunbing1   

  1. 1.College of Mechanical and Electrical Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing,210016
    2.Shanghai Spaceflight Precision Machinery Institute,Shanghai,201600
  • Online:2018-11-25 Published:2018-11-27

[车间调度]基于信息素的多Agent车间调度策略

陈鸣1;朱海华1;张泽群1;金永乔2;王盈聪1;唐敦兵1   

  1. 1.南京航空航天大学机电学院,南京,210016
    2.上海航天精密机械研究所,上海,201600
  • 基金资助:
    国家自然科学基金资助项目(51805253,U1637211);
    航空科学基金资助项目(20161652015)

Abstract: Aiming at the disadvantages of traditional multi-Agent methods based on contractual network protocol,such as single target optimization,large communication volumes,and poor global performance optimization,a multi-Agent dynamic scheduling strategy was proposed based on pheromone.The strategy achieved indirect negotiation between Agents by pheromone, reduced traffic and realied global multi-objective optimization.In addition,the task allocation stages and buffer job selection stages were optimized at the same time.The independent setup time was taken into account,which was more practical and further improved the overall system optimization effectiveness.Finally,an example simulation was used to verifiy the efficiency of the strategy.

Key words: pheromone, multi-Agent system, multi-objective optimization, independent setup time

摘要: 针对基于合同网协议的传统多Agent方法优化目标单一、通信量大、全局性能优化效果差的缺点,提出了基于信息素的多Agent动态调度策略。该策略通过信息素实现了Agent间的间接协商,减少了通信量,实现了全局的多目标优化。此外,采用该策略同时对生产任务分配阶段和缓冲区工件选择阶段的调度进行了优化,并且考虑了独立的调整时间,更加符合实际且进一步提升了系统整体的优化效果。最后,通过实例仿真验证了上述策略的效率。

关键词: 信息素, 多Agent系统, 多目标优化, 独立调整时间

CLC Number: