混合品种装配线平衡问题的一种混合搜索机制的蚁群算法_张则强.pdf

第 45 卷第 5 期 机 械 工 程 学 报 Vol.45 No. 5 2009 年 5 月 JOURNAL OF MECHANICAL ENGINEERING May 2009 DOI ：10.3901/JME.2009.05.095 混合品种装配线平衡问题的一种混合 搜索机制的蚁群算法* 张则强 程文明 钟 斌 王金诺 (西南交通大学机械工程研究所 成都 610031) 摘要：为有效求解混合品种装配线平衡问题，通过组合不同品种的优先顺序图，将混合品种装配线转化为单一品种的装配线 形式。提出了一种带信息素总合规则的混合搜索机制的蚁群算法，通过在任务和任务分配序列的位置之间释放信息素、采用 信息素总合规则以进行更有效的信息素累积，构造了综合考虑利用、探索和随机搜索的混合搜索机制，考虑了局部信息素更 新和全局信息素更新。为提高搜索效率，以协同考虑装配任务作业时间和后续任务数的分级位置权重作为蚁群算法的启发式 信息。最后通过实例验证，说明了算法的有效性。 关键词：混合品种装配线平衡 蚁群算法 启发式方法 中图分类号：TH165 TP301.6 Hybrid Behavior Ant Colony Optimization for Mixed-model Assembly Line Balancing Problem ZHANG Zeqiang CHENG Wenming ZHONG Bin WANG Jinnuo (Research Institute of Mechanical Engineering, Southwest Jiaotong University, Chengdu 610031) Abstract: For the purpose of solving mixed-model assembly line balancing problem (MMALBP), MMALBP is transformed into a single-model assembly line balancing problem with a combined precedence diagram. A hybrid behavior ant colony optimization with pheromone summation rules for MMALBP is proposed. The proposed algorithm makes use of the trail information which is deposited between the task and the task selected position, and pheromone summation rules are adopted. Hybrid search mechanism, which comprehensively considers utilization, exploration and random search, are adopted. Global pheromone trail update and local pheromone trail update are considered. The heuristic information is set to the position weight for tasks of MMALBP, which collaboratively considers the operation time of assembly task and the number of follow-up tasks. Finally, example ve