Hardware Implementation of a RBF Neural Network Controller with a DSP 2812 and an FPGA for Controlling Nonlinear Systems

This paper presents the hardware implementation of the neural network controller for controlling nonlinear systems. The neural network controller is implemented on the digital signal processing (DSP) chip and a field programmable gate array (FPGA) chip. The DSP 2812 controller board has been developed for controlling motors. Combining the DSP and the FPGA yields the neural network controller. The reference compensation technique (RCT) as a neural network learning algorithm is implemented. Experimental studies of balancing the angle and controlling the cart of the inverted pendulum system have been conducted to confirm the performance of the hardware implementation of the neural controller.

[1]  Mitsuo Kawato,et al.  Feedback-error-learning neural network for trajectory control of a robotic manipulator , 1988, Neural Networks.

[2]  T. Lahdhiri,et al.  Cart-pendulum balancing problem using fuzzy logic control , 1994, Proceedings of SOUTHEASTCON '94.

[3]  Zexiang Li,et al.  Stabilization of a 2-DOF spherical pendulum on X-Y table , 2000, Proceedings of the 2000. IEEE International Conference on Control Applications. Conference Proceedings (Cat. No.00CH37162).

[4]  Rogelio Lozano,et al.  Global stabilization of the cart-pendulum system using saturation functions , 2003, 42nd IEEE International Conference on Decision and Control (IEEE Cat. No.03CH37475).

[5]  Mario E. Magaña,et al.  Fuzzy-logic control of an inverted pendulum with vision feedback , 1998 .

[6]  Seul Jung,et al.  Decoupled Neural Network Reference Compensation Technique for a PD Controlled Two Degrees-of-Freedom Inverted Pendulum , 2004 .

[7]  G. Young,et al.  An interdisciplinary control systems laboratory , 1996, Proceeding of the 1996 IEEE International Conference on Control Applications IEEE International Conference on Control Applications held together with IEEE International Symposium on Intelligent Contro.

[8]  Seul Jung,et al.  Balancing and position tracking control of an inverted pendulum on a x-y plane using decentralized neural networks , 2003, Proceedings 2003 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM 2003).

[9]  Rong-Jong Wai,et al.  Development of adaptive sliding-mode control for nonlinear dual-axis inverted-pendulum system , 2003, Proceedings 2003 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM 2003).