基于C空间分解—路标法的机械手路径规划  

Path planning for a manipulator based on the decomposition of C space and the roadmap method

作　　者：周芳[1] 朱齐丹[1] 张智[1] 赵国良[1] 

机构地区：[1]哈尔滨工程大学自动化学院,黑龙江哈尔滨150001

出　　处：《哈尔滨工程大学学报》2009年第5期531-535,共5页Journal of Harbin Engineering University

基　　金：国家人事部出国留学人员基金资助项目(20001024A2)

摘　　要：针对核反应堆检修用多关节机械手的路径规划问题,提出了一种基于C空间分解-路标法的路径规划算法.由于该机械手C空间维数较高,因此将整个C空间分解成2组低维子空间,分别基于子空间建立路标图,并扩充成整个C空间内的路标图,从而将路标图的搜索维数由六维转化为三维,降低了路标图的搜索空间.同时为进一步提高路标图的质量,采用自适应变步长法来选取路标节点,根据前两时刻机械手各关节点距障碍物距离的变化情况来决定下一路标点的选择.仿真结果表明该方法不仅缩短了在线实时规划的时间,而且提高了路径规划的成功率.To deal with the problem of collision-free path planning for a multi-joint manipulator used to repair nuclear reactors, a roadmap method based on C space decomposition was developed. Since the C space dimensions of this manipulator are very high, the whole C space was decomposed into two lower dimensional subspaces, the road- map was built in the subspaces, then it was extended to the roadmap of the whole C space. So a search in a six - dimensional space was transformed into a search in three-dimensional space, so the searching range was greatly re- duced. In order to further improve the quality of the roadmap, an adaptive variable-step method was proposed to choose roadmap points, with change information from the last two minimum distances between the manipulator joints and obstacles deciding the next posture point. The simulation results showed that the approach reduces search time and also improves the success rate of path planning.

关 键 词：机械手 路径规划 C空间 路标法 自适应变步长 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]