Skip to Main Content
A functional-linked cerebellar model neural network (FCMNN) equipped with sine-cosine perturbed Gaussian basis functions to online approximate an unknown nonlinear term in the system dynamics of a DC motor driver is proposed in this study. The sine-cosine perturbation in the Gaussian basis functions possessing the ability of handling rule uncertainties is quite useful for real-time applications. Then, an intelligent total sliding-mode control (ITSMC) system that is composed of a computation controller and a robust compensator is proposed. The computation controller including an FCMNN approximator is the main controller and the robust compensator is designed to eliminate the effect of the approximation error introduced by the FCMNN approximator upon system stability. The online parameter adaptation laws are derived based on a Lyapunov function so that the L2 tracking performance can be guaranteed. To reduce the parameter overtraining problem, a dead-zone parameter modification scheme is adopted so that the parameter tuning process will be stopped when a tracking index is smaller than a pre-specified threshold. Finally, the proposed ITSMC system is implemented on a 32-bit microcontroller for possible low-cost and high-performance industrial applications. The experimental results show that the ITSMC system can achieve favourable tracking performance and is robust against parameter variations in the plant.