Adaptive functional-link-based neural fuzzy controller design for a DC gear motor driver

This paper proposes an adaptive functional-link-based neural fuzzy control (AFNFC) system based on complementary sliding-mode approach. The proposed AFNFC system is composed of a neural controller and a robust compensator. The neural controller uses a functional-link-based neural fuzzy system (FNFS) to approximate an ideal complementary sliding-mode controller, and the robust compensator is designed to eliminate the effect of the approximation error between a neural controller and an ideal complementary sliding-mode controller. Since the consequent part of the fuzzy rules in FNFS uses an orthogonal Hermite polynomial-based non-linear combination, the proposed FNFS can approximate an ideal complementary sliding-mode controller with good learning accuracy. Finally, to enhance the control performance of the proposed AFNFC system, a 32-bit ×86-microprocessor is adopted for the implementation of the proposed control system. A comparison among the fuzzy sliding-mode control, the intelligent-complementary sliding-mode control, the supervisory fuzzy neural network control, and the proposed AFNFC system is made. The experimental results demonstrate that the proposed AFNFC system can achieve favorable tracking performance and robust with regard to parameter variations for a DC gear motor driver.