Control experiment of a wheeled drive mobile pendulum using neural network

Abstract

In this paper, control of a wheeled drive mobile pendulum system is presented. The inverted pendulum mounted on the wheeled drive mobile robot is controlled by neural network. Neural network learning algorithm is embedded on a DSP board and controls the angle of the pendulum and the position of the mobile robot along with PID controllers. Uncertainties in system dynamics are compensated by neural network in on-line fashion. Experimental results show that the performance of balancing of the pendulum and position tracking of the mobile robot is successful.