摘要
根据车间调度的动态复杂性,本文提出了一种将多智能体(Multi-Agent)技术应用于设备调度的模型.整个调度系统建立于任务Agent,资源Agent和协调Agent基础上,通过招标——投标机制及奖罚制度完成车间调度,调度过程的优化通过一个综合指标来实现.经过仿真实例验证,本系统很好的实现了多目标优化,并且保证了整个系统朝着全局最优的方向发展.
According to thecomplicated Shop-scheduling, Multi-agent architectureisbuilt in the scheduling model. The schedulingsystem isbasedonthree kinds of agents as task agent, resourceagentand harmonizing agent. The architecture can distribute the task to the best resource agent by the mechanism of inviting public bidding and the mechanism of reward- punishing, with the most importantparameter. We callthe parameter as synthesisindex, witch makes upofseveral para- meters. It can guarantee the system optimize.
出处
《河北工业大学学报》
CAS
2005年第2期105-109,共5页
Journal of Hebei University of Technology
基金
天津市教委科研基金(041107)
关键词
多智能体
招标-投标
奖罚制度
综合指标
multi-agent
invite public bidding
reward-punish
synthesis index
