Adaptive RBFNN‐based predictive control for the nanopositioning of an electrostatic MEMS actuator

Abstract

AbstractElectrostatic micro electro mechanical system (MEMS) as an actuator is one of the most widely used actuators in micro dimension systems. Due to the high nonlinear characteristics in the micro actuator's model, nanopositioning has been a challenging problem for researchers. In addition, the system is subject to uncertainties associated with the system's unknown parameters and un‐modelled dynamics. Thus, to overcome these problems and achieve precise positioning a robust non‐linear control method is required. In this research, an optimal voltage control algorithm is initially designed utilizing the nonlinear predictive control approach. The Taylor expansion sequence is used to estimate the final position of the parallel plate actuator. To approximate unknown disturbances, a radial basis function neural network (RBFNN) has been designed. The estimations made were used in updating the control input in the previous step, and the obtained results showed an increase in the robustness of the designed controller. The stability of the closed‐loop system is also analyzed. Finally, a comparison is accomplished between the potential of the proposed controller and that of the feedback linearization and sliding mode controllers. The obtained results indicate better performance and robustness of the suggested controller.