Dynamic Global PID Sliding Mode Control for MEMS Gyroscope Using Adaptive Neural Controller

Abstract

This paper derived a dynamic global proportional integral derivative (PID) sliding mode control based on adaptive radial basis function (RBF) neural controller for a micro electromechanical systems (MEMS) gyroscope. This approach gives a new dynamic global PID sliding mode manifold, which not only enables system trajectory to run on the global sliding mode surface at the start point more quickly and eliminate the reaching phase of the conventional sliding mode control, but also restrains the steady-state error and reduces the chattering via a dynamic PID sliding surface. Meanwhile, a RBF neural network (NN) system is employed to estimate the lumped uncertainty and eliminate the chattering phenomenon. Additionally, adaptive laws and dynamic global PID sliding control gains that ensure system stability in a Lyapunov sense are proposed, together with the techniques for deciding which basis function should be selected. Finally, the effectiveness of RBFNN dynamic global PID sliding mode control method is demonstrated.