摘要
分析多种类型零件投放车间时根据零件工艺路线特点，在不同机床上选用适当的启发式调度规则，实现车间调度周期内的多规则混合调度（ＭＤＲ－ＭｉｘｅｄＤｉｓｐａｔｃｈｉｎｇＲｕｌｅ）的方法。实验结果表明，针对具体加工任务和加工资源特点，通过有限步的仿真搜索所形成的多规则混合调度，算法性能比通常所采用的针对所有机床的单规则调度算法有明显提高。该方法已应用于作者开发的车间控制器中。
We extended the method of Ishii et al[1] for FMS(Flexible Manufacturing System)scheduling to the more general case of job shop scheduling,where,unlike FMS, the flow of workpieces is more flexible.We developed a knowledge -based job shop scheduling algorithm as the shop controller of an aircraft factory.This algorithm includes two important strategies:(1) Unlike Ishii et al,we determine the bottleneck machine by that total time needed by it to machine all the workpieces in the job queue is the highest among all machines.(2) We, after ascertaining that Ishii et al gave no clues at all,deem that the worst workpiece should be that one that requires the largest flow time.Figs.2 and 3 give the results of our stochastic simulation experiments with SIMAN language model.Figs.2 and 3 show the variations of MFT and (MFT+MAL) respectively with strategy searching steps.Figs.2 and 3 show clearly that our mixed dispatching rule method is more effective than single rule method in dealing with this characteristics of job scheduling.
出处
《西北工业大学学报》
EI
CAS
CSCD
北大核心
1996年第4期627-632,共6页
Journal of Northwestern Polytechnical University
基金
国家863CIMS主题资助
国家自然科学基金
