A New Algorithm for Solving Inverse Kinematics of Robot Based on BP and RBF Neural Network

Abstract

A parallel neural network algorithm based on BP and RBF neural network for solving inverse kinematics of robot is proposed in this paper. Concrete steps of this method and related matters that should be noticed are presented. BP network is trained by LM algorithm and RBF network increases radial basis neurons automatically. The simulation results of PUMA560 show that this algorithm is simple and reliable, making the error of the whole system become smaller. In addition, the algorithm effectively solves the problem of inverse kinematics and overcomes the defects of traditional methods for solving inverse kinematics, such as large amount of calculation, slow convergence rate and low accuracy.