Applications of neural networks for control of a double inverted pendulum

Abstract

A study regarding the swing-up and stabilization problem of a double pendulum on a cart is presented. Current methods of feed forward control, energy based control, and passivity based control are used to select a design for the swing-up controller. The stabilization is done using an LQR controller. The nonlinear controller is analyzed with respect to being approximated by a neural network [1].