基于局部搜索NSGA-Ⅱ算法的机械产品分组选择装配方法

Grouping Selective Assembly Method for Mechanical Products Based on Local Search Strategy for NSGA-Ⅱ

针对批量机械产品多目标要求下的分组选择装配问题,提出一种基于局部搜索的第二代非支配排序遗传算法(NSGA-Ⅱ)的分组选择装配方法。以装配精度和装配成功率为优化目标,构建基于质量损失函数和尺寸偏差的综合选择装配模型。采用浮点数和整数相结合的编码方式来描述分组选择装配方案,为获得分组选择装配方案中每个产品的封闭环实际尺寸,设计了一种解码方式。为了提高算法局部搜索能力,利用NSGA-Ⅱ混合模拟退火算法对分组选择装配问题进行求解,将成绩标量函数作为个体评价规则,实现了多目标要求下选择装配分组方案的优化。以某深沟球轴承的装配为例,验证了该方法的可行性和有效性。

Aiming at the grouping selective assembly problems under multi-objective requirements for mass-produced mechanical products,a method was proposed based on local search strategy for non-dominated sorting genetic algorithmⅡ(NSGA-Ⅱ).Taking the assembly precision and assembly yield as optimization objectives,a comprehensive selective assembly model was established based on quality loss function and dimensional deviation.A combination of floating-point coding and integer coding was used to describe grouping selective assembly scheme.A decoding method was designed to obtain the actual size of the closed loop of each products in grouping selective assembly scheme.The achievement scalarizing function was taken as evaluation rule of the individuals,and the grouping selective assembly problem was solved by NSGA-Ⅱhybrid simulated annealing algorithm,which improved the local search ability and realized the optimization of multi-objective assembly.The assembly of a deep groove ball bearing was used to verify the feasibility and effectiveness of the method herein.