Neutral network compensator based cascade control for ball and beam system

This paper presents a neural network compensator based cascade control dynamics for the nonlinear control of the ball and beam system including actuator dynamics. Firstly, the cascade control structure divides the ball and beam model into two lower order discrete time models and according to the two discrete time models design the inner closed loop controller and the outer closed loop controller. Furthermore, this control scheme can deal with problems that other control methods can hardly achieve satisfied control performance due to the actuator dynamics of the ball and beam system. In addition, for the inner closed loop controller, radial basis function (RBF) network is integrated to compensate the nonlinearity in addition to traditional PD control. Finally, the proposed control strategy is based on discrete time and can easily be applied to industrial application. Experimental results demonstrate that the addressed control strategy can easily be implemented and achieve perfect control precision.