基于RBF神经网络的码垛机器人轨迹优化  

Trajectory Optimization of Palletizing Robot Based on RBF Neural Network

作　　者：刘一均[1] Liu yijun(SiChuan Water Conservancy Vocational College,ChengDu 611231,China)

机构地区：[1]四川水利职业技术学院,成都611231

出　　处：《包装与食品机械》2019年第2期62-65,共4页Packaging and Food Machinery

摘　　要：为了提高码垛机器人笛卡尔空间轨迹规划算法的收敛性和稳定性,提升机器人运行轨迹圆滑度。首先从运动学角度出发,建立了码垛机器人运动学模型。针对码垛机器人笛卡尔空间轨迹规划提出了一种基于RBF神经网络的码垛机器人轨迹优化算法。仿真结果表明,经过RBF神经网络轨迹优化使得在拟合的曲线更加接近实际的曲线,减小了逼近误差,提高了规划的准确性。该方法能够有效提升码垛机器人运动轨迹圆滑度,保证码垛机器人能够按照预定轨迹路径进行运动,有效实现了机器人笛卡尔空间轨迹规划。In order to improve the convergence and stability of the Cartesian space trajectory planning algorithm for the palletizing robot and the smoothness of the robot's running trajectory,firstly,the kinematics model of the palletizing robot was established based on kinematics. A trajectory optimization algorithm for the palletizing robot based on RBF neural network was proposed for Cartesian space trajectory planning of the palletizing robot.The simulation results show that the RBF neural network trajectory optimization made the fitted curve closer to the actual curve,which reduces the approximation error and improves the accuracy of the planning. This method can effectively improve the smoothness of the motion trajectory of the palletizing robot,ensure that the palletizing robot can move according to the predetermined trajectory path,thereby effectively realizing the trajectory planning of the robot in Cartesian space.

关 键 词：码垛机器人 轨迹规划 RBF神经网络 仿真 

分 类 号：TB485.3[一般工业技术—包装工程]