Research on Inverse Kinematics of Manipulator Based on Neural Network

Abstract

The inverse kinematics solution of manipulator is an important part of robot dynamics, trajectory planning and trajectory tracking control. This paper presents an inverse kinematics solution method of manipulator in which the RBF network is used. This method overcomes many defects of traditional inverse kinematics solution methods, such as low precision and slow convergence speed, and provides a new solution for the inverse kinematics of manipulator. Experimental results show that the RBF network has compact topological structure, separation and learning of structural parameters, fast convergence speed and global approximation ability, which fundamentally solves the local optimization problem of BP network. The scheme designed in this paper is simple and practical, and the trained parallel neural network has fast convergence speed and high precision, which can meet the requirements of high-speed inverse kinematics in the real-time control of manipulator.