Adaptive neural sliding control of MEMS gyroscope with robust feedback compensator

Abstract

A novel robust neural sliding control using the radial basis function (RBF) neural network for a micro-electro-mechanical system (MEMS) z-axis gyroscope is proposed in this paper. A robust feedback compensator is incorporated into the normal adaptive sliding control in order to eliminate the effect of model uncertainties and external disturbances, and improve the robustness of the whole control system. An adaptive RBF is also adopted to learn online the upper bound of external disturbances, relaxing the requirement of prior knowledge of the upper bound of model uncertainties. By using the proposed neural control, satisfactory dynamic characteristics, strong robustness with respect to external disturbances and parameter uncertainties and fast convergence of tracking errors to zero can be all obtained. In addition, the proposed adaptive neural controller can estimate the angular velocity and all the gyroscope parameters including damping and stiffness coefficients. All the adaptive laws are derived in the same Lyapunov framework, so that the stability of the closed-loop system can be guaranteed. Numerical simulations for MEMS gyroscope are implemented to verify the effectiveness of the proposed control scheme.