流水车间作业提前/拖期调度问题研究  

Earliness/tardiness scheduling problem in flow shop

作　　者：李素粉[1] 朱云龙[1] 

机构地区：[1]中国科学院沈阳自动化研究所,辽宁沈阳110016

出　　处：《计算机集成制造系统》2006年第8期1235-1240,共6页Computer Integrated Manufacturing Systems

基　　金：国家自然科学基金资助项目(70431003);国家863/CIMS主题资助项目(2003AA412210);国家重点基础研究规划资助项目(2002CB312204-03)~~

摘　　要：在非正规性能指标提前/拖期调度问题中,工件的加工顺序和每个加工活动的开始时刻都属于需要优化的变量,增加了求解的难度。针对这一问题,提出了采用分层调度模式求解流水车间提前/拖期调度问题的联合算法。首先,采用遗传算法对加工顺序进行寻优;其次,在给定调度序列的情况下采用启发式算法对加工开始时刻进行优化,制定插入机器空闲时段的策略,确定何时插入空闲时段和空闲时段的大小,即在给定顺序下确定工件加工活动的开始时刻,以满足在加工完所有工件后,使提前惩罚费用与拖期惩罚费用之和最小。数值计算结果证明了该联合算法的有效性。As a kind of unconventional scheduling problem, the earliness and tardiness performance function is not a monotony increasing relation on completion time of every job. Both job sequence and starting time of each operation must be optimized, which increased difficulties in solving this kind of problem. In order to achieve better schedulers for Earliness/Tardiness （ET） problems, a hierarchy scheduling paradigm was introduced and a jointed algorithm to solve problems based on genetic algorithm and heuristic algorithm was advanced. Firstly, the genetic algorithm was used to determine scheduling sequence preference. Secondly, a kind of heuristic algorithm was put forward to adjust the starting time for determined scheduler. Strategy for insertion of equipment idle time was established to define insertion time and idle time length. Performance was measured by the minimization of the weighed sum of E/T penalties of jobs. Numerical calculation results were used to demonstrate feasibility and effectiveness of the proposed method by comparing with genetic algorithm method.

关 键 词：流水车间 提前/拖期调度 遗传算法 启发式算法 

分 类 号：TB49[一般工业技术]