摘要
混流车间调度问题有很强的工程背景,一直是调度领域的研究热点。针对简单遗传算法在求解混流车间调度问题时存在的早熟收敛和易陷入局部极值点的现象,提出了多对染色体遗传算法。多对染色体提供了保留低适应度个体中的有用的染色体的功能,这一染色体构成最优解的部分染色体,以增强算法的搜索能力,提高搜索精度;连锁互换交叉算子可以增加个体的多样性,扩展解的搜索空间,进而增强算法的抗早熟能力。仿真实验表明,多对染色体遗传算法比简单遗传算法提高了全局收敛性能,是解决混流车间调度问题的有效方法。
Hybrid Flow Shop Scheduling Problem is very common in industry, and it is a hotspot in the study of scheduling problems. Genetic algorithm has been applied to HFSP, and it bears the common defect of premature convergence. A novel multi - pairs of chromosomes genetic algorithm （MCGA） is proposed. Because a certain chromosome of an individual with lower fitness to form one chromosome of the optimal solution is reserved, the search capability of MCGA is enhanced. Furthermore, a chain and interchange crossover improves the diversity of the population and avoids the prematurity of the population. The result of simulation test illustrates that MCGA compared with SGA improves the performance of global convergence, so it is an effective method for HFSP.
出处
《计算机仿真》
CSCD
2006年第2期157-160,共4页
Computer Simulation
基金
教育部高校骨干教师资助项目(0X104110)
