• •    

基于六维线性插值的六自由度机械臂逆运动学方程求解研究

周锋1,林楠2,陈小平1   

  1. 1. 中国科学技术大学多智能体系统实验室
    2. 中国科学技术大学 多智能体系统实验室
  • 收稿日期:2017-06-16 修回日期:2017-08-29 发布日期:2017-08-29
  • 通讯作者: 周锋

Research on inverse kinematics equation of six DOF Manipulator Based on six dimensional linear interpolation

  • Received:2017-06-16 Revised:2017-08-29 Online:2017-08-29
  • Contact: Feng ZHOU

摘要: 针对一般结构的六自由度机械臂逆运动学方程求解困难的问题。首先拓展一维、二维线性插值理论,提出六维线性插值理论,使用七个相邻的非线性相关的节点组成超体,利用这七个节点会得到六元一次线性预测函数,进而可以进行插值和反插值运算。使用Matlab按照正运动学方程产生一百万组经验数据,并对目标位姿进行反向插值,迭代预测六个关节角。实验结果表明,相比径向基网络,六维线性反插值法能够更快更准的逼近目标位姿。论文提出的方法是基于数据的算法,避免了复杂的理论,满足机器人日常应用的要求。

关键词: 机器人, 六自由度机械臂, 运动学, 逆运动学, 六维线性插值

Abstract: Abstract: Concern the problem of solving the inverse kinematics equation of six degree of freedom manipulator is discussed, a method six dimensional linear interpolation theory was proposed through expand the one-dimensional and two-dimensional linear interpolation theory. Super body composition using seven adjacent nonlinear related nodes, using the seven node will get six yuan for a linear prediction function, which can be used to interpolation and inverse interpolation. Matlab is used to generate one million sets of empirical data according to the forward kinematics equation, and the target position is inverted interpolated, and the six joint angles are predicted iteratively. Experimental results show that, compared with radial basis networks, the six dimensional linear inverse interpolation method can approach the target pose faster and more accurately. The method presented in this paper is a data based algorithm that avoids complex theory and meets the requirements of robot everyday applications.

Key words: Robot, 6-DOF manipulator, Kinematics, Inverse kinematics, Six dimensional linear interpolation

中图分类号: