基于配合约束的装配序列规划的遗传算法研究

Genetic algorithm for assembly sequences planning based on assembly constraint

为提高装配序列规划问题的寻优效率,提出了一种将装配约束条件和遗传算法相结合的求解方法。对配合约束进行了分类并给出每种类型特征尺寸的定义。在此基础上,引入特征标签的概念,用于描述零件的装配特征信息和配合约束关系。根据零件间特征标签的相似程度设计了可装配度函数,以度量装配的可行程度,并以该函数为基础构建了遗传算法的适应度函数。另外,为减少无用个体的数目,提出了基于配合约束的子代个体的保留策略,并以此为指导进行子代的有效性检验。最后,通过实例说明了该算法的有效性。

To improve the efficiency of assembly sequences planning, a new approach based on assembly constraints and genetic algorithm was proposed. Assembly constraints were classified and the characteristic dimension of each kind was defined. The concept of Feature Mark （FM） was introduced to describe assembly feature and assembly constraint. Assembly capability function was designed to measure the feasibility of assembly based on the similarity degree between FMs of parts. Then, fitness function was built according to assemble ability function. Furthermore, reserve strategy for descendants was proposed to reduce the number of useless descendants. The reserve strategy was also used to check the validity of descendants. Finally, the validity and feasibility of the above-mentioned algorithm was proved by an application example.