基于RBF神经网络的ABB-IRB460码垛机器人逆运动学算法

Inverse Kinematics Algorithm of ABB-IRB460 Palletizing Robot Based on BRP Neural Network

为解决传统逆运动学求解速度慢和精度低等问题,以流行的ABB-IRB460码垛机器人为研究对象,根据其物理构造和运动特点,将RBF神经网络应用于该型机器人的运动学逆解问题中,同时采用基于K-means++的聚类算法对RBF网络进行训练。实验结果显示:所提方法能提高网络训练速度,并降低网络输出误差,是一种有效的ABB-IRB460码垛机器人的运动学逆解方法,且求解精度较高,实时性能较好。

The traditional solution of inverse kinematics has lower accuracy and speed. To solve theproblem, ABB-IRB460 palletizing robot is analyzed in this paper, and according to its physical structureand movement characteristics, RBF neural network is introduced to the inverse kinematics solution of therobot. The K-means ＋＋ clustering algorithm is used to train the RBF neural network at the same time.The simulation results show that the proposed algorithm can improve the accuracy of network output andthe speed of Network training. It’s an effective solution of the ABB-IRB460 inverse kinematics, and hashigh accuracy and good real-time performance.