摘要
根据复杂曲面的几何特性和拓扑结构,对曲面进行分片处理,在每一片上进行喷涂机器人喷枪路径的规划。将整个曲面上的喷枪路径顺序组合问题看成乡村邮递员问题(ORPP),并采用哈密尔顿图形法表示ORPP,用遗传算法进行求解。个体编码采用代表哈密尔顿图中顶点的实数编码和代表每片上路径方向的编码相结合的方法,可实现对问题空间的全局快速寻优。最后以某品牌汽车车身后盖为喷涂对象进行仿真实验,实验结果验证了所提算法的有效性。
A complex curved surface has to be divided into several patches and the path planning for each patch was performed. The path integration problem can be modeled as an open rural postman problem (ORPP), and a graph transformation method by using a Hamiltonian graph represented all the required edges in the ORPP. A genetic algorithm was advanced to solve the ORPP. The chromosome of GA made use of both numbers of the vertexes in the Hamiltonian graph and numbers of the edges in the original graph, which was a new encoding method and was able to search optimal solutions within the entire fields. Simulations were carried out on automotive body parts and the results validate the proposed algorithm.
出处
《中国机械工程》
EI
CAS
CSCD
北大核心
2008年第7期777-779,792,共4页
China Mechanical Engineering
基金
江苏省高技术研究发展计划资助项目(BG2006023)
关键词
喷涂机器人
路径规划
遗传算法
路径组合
spray painting robot
path planning
genetic algorithm
path integration
作者简介
赵德安,男,1956年生。江苏大学电气信息工程学院教授、博士研究生导师。主要研究方向为机器人控制、智能控制、计算机控制。发表论文60余篇。
陈伟,男,1981年生。江苏科技大学电子信息学院助教。
汤养,女,1980年生。江苏大学理学院讲师。