摘要
采用遗传算法对一种关节型模块化机器人系统面向任务的拓扑构型优化设计问题进行了研究。对任意机械臂拓扑构型组成模块的类型、参数、数量及模块间连接顺序和装配方位等信息进行了二进制编码。按照性质不同将设计要求划分为刚性要求、柔性要求及刚性+柔性要求,对不满足刚性要求的拓扑结构通过过滤器处理以提高效率,对其他拓扑结构进行构型评价并以评价结果的加权和的形式建立构型评价函数。以自由度最小为主要优化目标,结合遍历法和遗传算法给出一种拓扑构型优化设计方法。构型优化实验验证了该方法的有效性。
A study of task-oriented topology configuration optimization for a joint type modular robot system was carried out based on genetic algorithm. The topology configuration information such as the module types, parameters, quantity, assembling sequence and orientations of any given manipulator were represented by binary coding. Design requirements were divided into hard ones, soft ones and hard＋soft ones according to their properties. The topology structure which did not meet the ＂hard＂ requirements would be processed by filters to improve the computational efficiency. Meanwhile, the others were evaluated and the configuration evaluation function was constructed based on the weighted sum of the evaluation results. Taking the minimized degree-of-freedom as the main optimization objec tive,a topology architecture optimization design method was given based on the combination of traversal method and genetic Mgorithm. The effectiveness of the method was verified through an arehitecture optimization experiment
出处
《中国机械工程》
EI
CAS
CSCD
北大核心
2014年第12期1574-1580,共7页
China Mechanical Engineering
基金
国家自然科学基金资助项目(60905048)
机器人学国家重点实验室自主课题(RLZ200802)
关键词
模块化机器人
可重构
构型优化
遗传算法
modular robot
reconfigurable
configuration optimization
genetic algorithm
