应用于贴片机贴装顺序优化的遗传算法的比较和改进  

Comparison and improvement of genetic algorithm applied to SMT machine for mounting sequential optimization

作　　者：罗爱玲[1] 龙绪明[1] 

机构地区：[1]西南交通大学,四川成都610031

出　　处：《现代电子技术》2014年第6期129-131,135,共4页Modern Electronics Technique

摘　　要：元件贴装顺序优化是决定贴片机生产效率的关键问题,传统的解决贴装顺序优化问题的方法有遗传算法,蚁群算法,SS(伞布搜索法)等。使用较多的还是遗传算法。遗传算法中包含选择算子、交叉算子、变异算子,且编程思想简单,但容易出现局部最优,过早收敛等情况。在此,通过对遗传算法在贴装顺序优化应用的结果比较找到一种更适合的遗传算法,使之拥有较快的收敛速度和全局优化性。Component placement sequence optimization is the key link which determines the production efficiency of SMTmachine. The traditional methods to complete the link are genetic algorithm,ant colony algorithm,SS（umbrella cloth searchmethod）and so on. Genetic algorithm（GA）is used more frequently. It contains the selection operator,crossover operator andmutation operator,and its programming thought is simple,but it＇s easy to result in a local optimum,premature convergence,etc. A more suitable genetic algorithm was found through the comparison of sequence optimization results of genetic algorithm,which has faster convergence speed and global optimization.

关 键 词：元件贴装顺序优化 贴片机 遗传算法 交叉算子 

分 类 号：TN919-34[电子电信—通信与信息系统] TP202.7[电子电信—信息与通信工程]