基于帝国主义和遗传混合算法的装配序列规划研究  

An Imperialist Competitive and Genetic Hybrid Algorithm for Assembly Sequence Planning

作　　者：曲倩雯[1] 杨志宏[1] 李娜[1] 

机构地区：[1]山东大学机械工程学院CADCAM研究所,山东济南250061

出　　处：《机械工程与自动化》2016年第5期7-9,12,共4页Mechanical Engineering & Automation

基　　金：国家自然科学基金资助项目(51375277);山东省科技发展计划资助项目(2013GGX10303)

摘　　要：根据复杂产品的序列规划特点,为提高求解效率,提出了面向序列规划的混合算法。利用遗传算法和帝国主义竞争算法各自的优点,将二者有机联合,以重定向次数、装配工具改变次数以及装配类型变化次数为约束条件来构造目标函数,提出最小装配成本概念。以一个包含8个零件的装配体实例进行MATLAB仿真试验,分析混合算法特性,并将混合算法与单独的帝国主义竞争算法和遗传算法进行比较。试验证明该混合算法在求解效率上明显优于单独的智能算法,且求得的序列更加符合实际的装配需求。According to the characteristics of assembly sequence planning of complex products,and in order to improve resolution efficiency,this paper presents a hybrid algorithm to solve assembly sequence planning.By combining the genetic algorithm and the imperialist competitive algorithm,and taking the alteration of assembly direction,the change of assembly tool and the change of assembly type as the constraint conditions,we build the fitness function,then we propose the concept of minimum assembly costs.The characteristics of the hybrid algorithm are analyzed through MATLAB simulation experiment of an assembly which contains eight parts.Meanwhile,the efficiency of the hybrid algorithm is compared with the imperialist competitive algorithm and genetic algorithm respectively.The simulation results indicate that the efficiency of the hybrid algorithm is much better than single algorithm,and the assembly sequence obtained conforms to the requirement of the real assembly.

关 键 词：装配序列规划 最小装配成本 混合算法 适应度函数 

分 类 号：TP391.7[自动化与计算机技术—计算机应用技术] TG95[自动化与计算机技术—计算机科学与技术]