摘要
提出了一种集成权重系数变化法和小生境技术的混合遗传算法,建立了包括时间、成本、交货期满意度和设备利用率在内的多目标优化模型。采用基于工序的编码方式和“间隙挤压法”活动化解码方法;遗传算子包括选择、交叉、变异3种类型;选择操作采用轮盘赌选择方式。为了保证解的收敛性和多样性,采用了精英保留策略和小生境技术。交叉操作采用线性次序交叉方式;变异操作采用互换操作变异方法。染色体的适应度是各个目标函数的随机加权和。仿真实验证明,提出的混合遗传算法可以有效解决柔性作业车间多目标调度优化问题。
To solve synchronization problem in assignment of machines to operations and tne scneduling of operattions on the assigned machines in Flexible Job shop Scheduling Problem （FJSP） with multi--objective, a hybrid genetic algorithm combining random weigh method with niche technology was proposed. Firstly, the multi--objective FJSP optimization model was built, where time, cost, delivery satisfaction and equipment utilization rate were all concerned. The operation--based encoding and an active scheduling decoding method were employed. There were three kinds of genetic operators, selection, crossover and variation. In the selection, the niche technology and the spirit strategy were integrated with the roulette selection operation to ensure the convergence and the diversity of the solution. Linear order crossover operation and reciprocated operation exchange were also used. The fitness of a chromosome was the sum total of all objectives with random weight. Finally,a simulation experiment was carried out to illustrate that the proposed method could solve multi--objective FJSP problem effectively.
出处
《计算机集成制造系统》
EI
CSCD
北大核心
2006年第5期731-736,共6页
Computer Integrated Manufacturing Systems
基金
国家863/CIMS主题资助项目(2003AA411110)
教育部博士点基金资助项目(20040699025)~~
关键词
柔性作业车间
多目标调度
遗传算法
权重系数变化法
小生境技术
flexible job shop
multi-- objective scheduling
genetic algorithm
random weigh method
niche technology
