Solving Complex Geometry Kinematics Problem with Function Approximation by Artificial Neural Network
Johns Hopkins University
Neural networks have been widely deployed to solve classification and regression problems, and, in recent years, there has been much interest in using neural networks for solving complicated problems. In this paper, a hybrid manipulator kinematics is studied, and an artificial deep neural network is used to solve the inverse kinematics problem. The noise resistance benefit of neural networks approach is explored.
Robot Kinematics, Artificial Neural Netwrok