SS7-7　Hybrid Scatter Search for dynamic job shop problem

◎Yedan Wu，Wei Weng，Shigeru Fujimura（Waseda University）

Job-shop Scheduling Problem (JSP) is one of the most well-known scheduling problems and one of the strongly NP-hard problems. Dynamic Job-shop Scheduling Problem (DJSP) includes machine breakdown and job arrival (event driven), which is much more complicated. To solve the problem, a new hybrid Scatter Search (SS) method is proposed, including modified variable neighborhood search (MVNS)for local search. The proposed method is compared with other algorithms such as MVNS, SS and FIFO. The experimental simulation shows that the hybrid method is quite effective at solving the event driven job shop scheduling problem.