MULTI-AGENT SCHEDULING FOR MACHINE BREAKDOWN AND RECOVERY
The problem considered in this research is the multi-machine scheduling problem with machine breakdown and recovery. The manufacturer factory environment consists of multiple stages, with multiple machines at each stage. The release time of each job is unknown. All jobs must go through all the stages in sequence to become a product. The processing time for each product on each machine of each stage is different. There exist delivery time between machines of two consecutive stages. Machines may suffer break down, and after a while the machines may be recovered and begin operation again. The objective of this research is to minimize the total earliness and tardiness penalties of all jobs. In this research, efforts have been made on online scheduling with new jobs arriving at the manufacturing system randomly. Multi-agent scheduling approaches are proposed to solve the problem. The multiple agents are featured by concurrent computing with distributed feedbacks. Each agent only knows limited information, but all agents work concurrently for the systematic objective. Computational simulation has been done to test the proposed approaches under disturbed system conditions such as machine breakdown and recovery. The result reveals that the proposed approach delivers competitive performance for the targeted problem even under unstable manufacturing situations.
Just-In-Time earliness and tardiness penalties agent-based feedback system dynamic scheduling.
Wei Weng Shigeru Fujimura
Graduate School of Information, Production and Systems, Waseda University, 2-7 Hibikino, Wakamatsu-k Graduate School of Information, Production and Systems, Waseda University,2-7 Hibikino, Wakamatsu-ku
国际会议
上海
英文
1-6
2009-08-02（万方平台首次上网日期，不代表论文的发表时间）