Neural Network Based Inverse Kinematics Solution for 6-R Robot Implement Using R Package Neuralnet

Abstract

R packages neuralnet is used to calculate the inverse kinematics of the ABB IRB7600-500 robot 6-revolute jointed robot arm. This robot arm is required to move from a position to the desired position with a minimum error in the workspace. The commonly methods are not sufficient, it can not solve the complex kinematics problems of the redundant robotic manipulators. To overcome such complicated things, ANNs are used nowadays. Firstly we analyzed the kinematics of 6-DOF robotic arm, secondly quintic polynomials was applied in 6-DOF robotic arm trajectory planning. The network is trained using the input data and the performance plot, training state and regression plots are observed. In this training, Random rule divides the data where 70% data are assigned to training set, 30% data to test set.at the same times, theta error were obtained acceptable.