Abstract: Aiming at the problem of assembly sequence planning,all the relevant operations of firefly algorithm which was always applied to optimize in continuous space were redefined,and the improved discrete firefly algorithm was proposed by improving the iteration rule of firefly algorithm.The experiment results of the improved and standard discrete firefly algorithm were compared,and the result showed that the improved discrete firefly algorithm was superior to the standard discrete firefly algorithm and genetic algorithm which was applied most frequently in assembly sequence planning field.In fitness function construction,the traditional research of assembly sequence planning was improved to create evaluation indicator of interference frequency between tools and parts,and the feasibility and effectiveness of this indicator was verified by application.

Key words: assembly sequence planning, firefly algorithm, fitness function

摘要： 针对装配序列规划问题的特点,重新定义了应用于连续空间优化的萤火虫算法的各种相关操作,并对萤火虫算法的迭代规则进行了改进,提出了面向装配序列规划问题的改进型离散萤火虫算法。将改进型离散萤火虫算法和标准型离散萤火虫算法的实验结果进行了比较,同时也将改进型离散萤火虫算法与在装配规划领域应用最广泛的遗传算法进行了比较,验证了改进型离散萤火虫算法的优越性。在适应度函数构造方面,对传统的装配序列规划研究进行了改进,提出了装配序列工具—零件干涉次数评价指标,并通过实例验证了该评价指标的可行性和合理性。

关键词: 装配序列规划, 萤火虫算法, 适应度函数