模块化主从机器人的运动学映射算法  

Kinematic Mapping Algorithms for Modular Master-Slave Robots

作　　者：苏满佳[1] 吴鸿敏[1] 谷世超[1] 黄晶[1] 管贻生[1] 刘冠峰[1] 

机构地区：[1]广东工业大学机电工程学院,广东广州510006

出　　处：《机器人》2015年第6期725-731,共7页Robot

基　　金：广东省自然科学基金重点项目(S2013020012797);国家自然科学基金委-广东省人民政府联合基金重点项目(U1401240);国家自然科学基金项目(51375095);国际科技合作专项项目(2015DFA11700)

摘　　要：以模块化主从机器人为研究对象,提出5种基本的运动学映射——PO3映射、O3映射、P3映射、p O3映射与Po3映射,和一种由5种基本映射组成映射序列的遥操作任务实现方法.由5种基本映射组成映射序列是一种通用的遥操作任务实现方法,能应用于绝大部分同构或异构的遥操作系统.此外,对由操作者、主机器人、从机器人和任务对象组成的全闭环控制回路作了分析,着重剖析操作者在控制回路中的正、负双面作用——高智能决策与误差引入.最后,以仿生攀爬机器人夹取球体、抓夹圆杆和方形杆为3种代表性任务,建模、仿真及验证了所提出的遥操作实现方法在操作者正、负双面作用下的可行性,实现的效果既反映了完成遥操作任务的往复过程,也展示了本映射方法的对单个任务的可行性和对多种任务的适应性.For modular master-slave robots, 5 basic kinematic mapping algorithms are proposed, including PO3, O3, P3, pO3 and Po3, and an implementation method for teleoperation task through mapping sequence consisting of those 5 basic mappings is presented. The presented method is commonly suitable for both isomorphic and isomerous master-slave teleoperation systems. Besides, a whole closed control loop consisting of the operator, master robot, slave robot and task object is analyzed, especially the positive and negative effects of the operator, i.e. high intelligent decision and error sources. Finally, the presented implementation method for teleoperation task is modeled and simulated through 3 classic tasks of climbing bio-robot, including grasping a ball, a cylinder and a cube. Hence, the feasibility of the method under positive and negative effects is verified. The implementation performance demonstrates the repeated process for completing teleoperation tasks, and show that the proposed method can perform a single task and is suitable for various kinds of tasks.

关 键 词：遥操作 模块化机器人 运动学映射 映射序列 操作者误差 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]