一种求解作业车间调度问题的文化遗传算法.docxVIP

一种求解作业车间调度问题的文化遗传算法 一种求解作业车间调度问题的文化遗传算法 一种求解作业车间调度问题的文化遗传算法) ) ) 王伟玲 李铁克 施灿涛 一种求解作业车间调度问题的文化遗传算法 王伟玲 李铁克 施灿涛 北京科技大学, 北京, 100083 摘要:针对传统遗传算法缺乏有效指导, 容易陷入局部极值的缺点, 提出了以一种采用种群空间和信仰空间的双层进化结构进行寻优的作业车间调度算法。该算法针对调度问题的特点, 以遗传算法为主群体空间, 利用优良调度方案的知识信息构成信仰空间。为充分利用父代个体的优良特征加速收敛, 算法采取不同的策略在主群体空间中指导遗传操作, 在选择操作中引入k 近邻法的思想进行动态学习, 在变异操作中通过选择合适的变异点进行邻域搜索变异。典型算例的仿真实验与分析表明, 算法在计算效率和求解质量上均具有较好的效果。 关键词:作业车间调度; 文化遗传算法; 邻域搜索变异; k 近邻法 中图分类号:T P301. 6 文章编号:1004) 132X(2021) 03) 0303) 07 An Effective Cultural Genetic Algorithm for Job Shop Scheduling Problem Wang Weiling Li Tieke Shi Cantao University of Science and T echnology Beijing, Beijing, 100083 Abstr act :Aiming at the disadvantages of traditional genetic algorithms that are lack of efficient guidance and easy to get into local extremum, this paper developed a double evolution frame popula 2tion space and belief space to solve job shop scheduling pr oblem. CGA was to extract the excellent in 2dividuals schema of the scheduling solution from the population space of genetic algorithm as the use 2ful knowledge to form belief space. A nd the belief space was utilized to guide the genetic operator of selection and mutation by two different ways respectively in order to make use of the characteristics of excellent individuals. k -near est neighbor method was introduced to do dynamic learning in selection operator, and do neighbor sear ch mutation in mutation operator at an appropriate position. T he differ 2ent sizes of the benchmark data taken from literature wer e used to analyze the efficiency of this algo 2r ithm. Exper imental results indicate that it outperforms current approaches in computational time and quality of the solutions. Key words :job shop scheduling; cultural genetic algorithm(CGA) ; neighborhood search mutation; k -near est neighbor method 车间作业调度问题已被证明属于典型的NP -hard 组合优化问题[1], 主要原因是:1解空间随着工件数和机器数非线性增加; o工件的各个工序之间存在机器顺序约束。调度问题的早期研究主要集中在开发有效的启发式规则, 但由于启发式规则自身的局限性, 研究者们开始关注