摘要
并联机器人是一种具有高载荷自重比的封闭式运动结构,针对并联机器人运动控制和NURBS轨迹问题进行了深入的研究,首先从并联机器人的逆运动学问题进行了解析方法的求解。其次,针对正运动学(FKP)在数学上是难以解决问题,提出了一种多层感知器进行反向传播学习的神经网络进行实时求解。再次,开发了基于NURBS的通用插补器,它可以处理任何类型的几何图形使得机器人运动轨迹平滑。最后利用实验验证了运动学和NURBS曲线求解并联机器人模型的正确性。该策略在少数迭代和很少执行时间内,位置和方向参数的精度分别接近0.01mm和0.01°,验证了算法的有效性和正确性。
Parallel robot is a closed motion structure with a high load to weight ratio.In this paper,the motion control of parallel robot and NURBS trajectory are studied in depth.Firstly,the inverse kinematics of parallel robot is solved by analytical method.Secondly,the forward kinematics(FKP)is difficult to solve mathematically,this paper proposes a multi-layer perceptron back-propagation learning neural network for real-time solution.Thirdly,a universal interpolator based on NURBS is developed,which can handle any type of geometric shapes to make the robot’s trajectory smooth.Finally,the correctness of kinematics and NURBS curves for solving parallel robot model is verified by experiments.The accuracy of position and direction parameters of this strategy is close to 0.01mm and 0.01°respectively in few iterations and few execution time,which verifies the effectiveness and correctness of the algorithm.
作者
张皓宇
刘晓伟
任川
赵彬
ZHANG Hao-yu;LIU Xiao-wei;REN Chuan;ZHAO Bin(Researcher Level Senior Engineer of Meteorological Information Center,Liaoning Shenyang110168,China;SIASUN Robot&Automation Co.,Ltd.,Liaoning Shenyang110168,China;School of Information Science&Engineering,Northeastern University,Liaoning Shenyang110819,China)
出处
《机械设计与制造》
北大核心
2021年第4期282-286,292,共6页
Machinery Design & Manufacture
基金
国家高技术研究发展863计划(No.2012AA041405)
沈阳市工业科技攻关项目(No.F12-010-2-00)。