摘要
针对柔性车间调度提出改进遗传算法,采用机器分配链和工序顺序链的双链结构编码;对机器分配链设计基于拟水平均匀设计的初始化方法,对相应的工序顺序链采用剩余时间最短的启发式初始化方法;采用新生策略改进新一代种群的生成;有针对性地对个体的瓶颈工序进行交叉操作;基于极限最优适应度值和当前最优适应度值对种群个体选择性解码.针对常用的典型算例进行多方面的实验计算,并对实验结果进行对比分析,验证了改进遗传算法的有效性.
An improved genetic algorithm was proposed for flexible job shop scheduling.Double chains structure with machines＇allocation chain and operations＇sequence chain was used to encode the chromosomes.Population＇s machine-allocation-chains were initialized with quasi-level uniform design,and their operation-sequence-chains were heuristically initialized with longest remaining processing time first rule.New population was produced with some new-born chromosomes.Population crossover was improved with bottleneck-operation oriented method.Chromosome＇s decoding was controlled with two rules based on the extreme fitness and the current optimal fitness.Multi-aspects case studies based on some typical benchmark examples from the literature were conducted,and the experimental results were compared.Results showed a quicker evolution speed and powerful optimizing capability.
出处
《浙江大学学报（工学版）》
EI
CAS
CSCD
北大核心
2012年第4期629-635,共7页
Journal of Zhejiang University：Engineering Science
基金
浙江省制造业信息化重大科技攻关项目(2006C11234)
国家"863"高技术研究发展计划资助项目(2011AA040601)
关键词
柔性车间调度
遗传算法
瓶颈工序
拟水平均匀设计
启发式初始化
flexible job shop scheduling
genetic algorithm
bottleneck operation
quasi-level uniform design
heuristic population initiation
