摘要
针对最小化最大完工时间的置换流水车间调度问题,提出一种将遗传算法与蚁群算法相结合的改进区块遗传算法。算法利用随机机制和改进反向学习机制相结合的方式产生初始解,以兼顾初始种群的多样性和质量。通过若干代简单遗传算法操作产生精英群体,借鉴蚁群算法中利用蚂蚁信息度浓度统计路径和节点信息的思想,对精英群体所携带信息进行统计分析并建立位置信息素矩阵和相依信息素矩阵,根据两矩阵挖掘区块并将区块与非区块组合形成染色体。将染色体进行切段与重组,以提高染色体的质量,使用二元竞赛法保留适应度较高的染色体。算法通过 Reeves实例和 Taillard实例进行测试,并将结果与其他算法进行比较,验证了该算法的有效性。
Targeting the permutation flow-shop scheduling problem that minimizes maximum completion times,an im-proved block genetic algorithm combined with an ant colony algorithm is proposed.The algorithm uses a random mech-anism and the improved opposition-based learning mechanism to generate the initial solution.It takes into account the diversity and quality of the initial population.Elite populations are generated through several generations of simple ge-netic algorithm operations.Based on the idea of using ant information density concentration’s statistical path and node information in the ant colony algorithm,the information carried by elite groups was counted.Position pheromone and dependent pheromone matrices were established.Mining blocks according to two matrices were developed and blocks were combined with non-blocks to form artificial chromosomes.Chromosomes were cut and recombined to increase chromosome quality.The binary race method was used to retain chromosomes with higher fitness.The algorithm was tested through the Reeves and Taillard instances,and the results were compared with other algorithms to verify effect-iveness of the algorithm.
作者
裴小兵
张春花
PEI Xiaobing;ZHANG Chunhua(School of Management,Tianjin University of Technology,Tianjin 300384,China)
出处
《智能系统学报》
CSCD
北大核心
2019年第3期541-550,共10页
CAAI Transactions on Intelligent Systems
基金
国家创新方法工作专项项目(2017IM060200)
天津市哲学社会科学规划项目(TJYY17-013)
关键词
生产调度
组合优化
遗传算法
蚁群优化算法
构建区块
人工染色体
production scheduling
combinatorial optimization
genetic algorithms
ant colony optimization
buildingblock
artificial chromosome
