面向第Ⅱ类装配线平衡问题的蚁群算法  

Ant colony optimization for type Ⅱ assembly line balancing problem

作　　者：郑巧仙[1,2] 李元香[1] 李明[3] 唐秋华[4] 鲁素丽[4] 

机构地区：[1]武汉大学计算机学院,湖北武汉430072 [2]湖北大学数学与计算机科学学院,湖北武汉430062 [3]武汉科技大学理学院,湖北武汉430065 [4]武汉科技大学机械自动化学院,湖北武汉430081

出　　处：《计算机集成制造系统》2012年第5期999-1005,共7页Computer Integrated Manufacturing Systems

基　　金：国家自然科学基金资助项目(50875190;61074036);武汉科技大学节能减排中心开放基金资助项目(B1019);冶金工业过程系统科学湖北省重点实验室开放基金资助项目(Y201109)~~

摘　　要：针对第Ⅱ类装配线平衡问题,提出一种基于可行装配序列的改进蚁群算法。算法基于可选操作集合的动态改变和工位作业时间优化目标的更新,给出操作分配至工位的分配准则。针对该问题的特点,提出工位和操作间的信息素、操作和操作间的信息素两种信息素。蚂蚁根据前者和启发式因素的权值为当前工位随机选择一项操作为该工位的首项操作,依据后者和启发式因素的权值为已选操作组合随机选择一项操作作为其组合操作。利用与经典测试算例的比较及工业实例的运行,验证了算法的正确性和工业应用优势。Aiming at the type Ⅱassembly line balancing problem（SALBP-2）,an improved ant colony optimization based on feasible assembly sequences was proposed.Based on the dynamic changes of candidate task set and the updating of station time optimization goal,a principle of assigning task to station was designed.By taking the characteristics of SALBP-2 into consideration,two pheromones were introduced into algorithm,which included station and task,task and those just among tasks.Ants selected the first task to station randomly according to the first pheromone and the weight of heuristic factors,and selected other tasks to station randomly according to the second pheromone and the weight of heuristic factors.Through the operation of industrial case and the comparison with classic test example,the validity and efficiency of the proposed algorithms was confirmed.

关 键 词：蚁群算法 装配线平衡 NP难问题 

分 类 号：TH166[机械工程—机械制造及自动化] TP18[自动化与计算机技术—控制理论与控制工程]