基于VPS-PSO算法的装配序列规划方法  

Assembly Sequence Planning Based on Various Population Strategy-Particle Swarm Optimization Algorithm

作　　者：刘冬[1] 张卫[1] 陆宝春[1] 

机构地区：[1]南京理工大学机械工程学院,南京210094

出　　处：《组合机床与自动化加工技术》2017年第2期30-33,共4页Modular Machine Tool & Automatic Manufacturing Technique

摘　　要：针对装配序列规划问题的特点,提出一种求解装配序列规划问题的变种群策略-粒子群优化(Various Population Strategy-Particle Swarm Optimization,VPS-PSO)算法。针对粒子群算法容易陷入局部最优的缺点,采用变种群策略,缩短进化停滞时间,提高粒子群算法进化效率,增强算法的寻优能力。并结合装配几何可行性、装配过程连续性、装配工具改变次数3个评价指标构建适应度函数,实现多目标优化。以经编机成圈传动机构装配序列规划实例验证VPS-PSO算法比较PSO算法具有更好的全局搜索能力。Aiming at the problem of assembly sequence planning,the Various Population Strategy- Particle Swarm Optimization( VSP-PSO) algorithm which is applied to solve the problem of the assembly sequence planning. To rise above the deficiency that PSO algorithm is easy to fall into local optimization,using the various population strategy,which is taken up to improve the optimizability of PSO algorithm to shorten the stagnancy of time in evolution,and enhance the ability of the algorithm optimization. Realizing the multiobjective optimization by designing the objective function based on the geometrical feasibility,the process continuity and assembly tools change times. An exemplification as the loop-forming transmission mechanism assembly application of the warp knitting machine demonstrates that global searching ability of VSP-PSO algorithm is more efficient compared with PSO algorithm.

关 键 词：装配序列规划 粒子群算法 变种群策略 多目标优化 

分 类 号：TH162[机械工程—机械制造及自动化] TG506[金属学及工艺—金属切削加工及机床]