Nonlinear Model Predictive Control for a Maglev Vehicle Regarding Magnetic Saturation and Guideway Irregularities

Abstract

In this work, a nonlinear magnet model including magnetic saturation is proposed with focus on control design for a magnetically levitated (Maglev) vehicle. It is shown that the model’s static characteristics corresponds with measurements obtained from a test bench, whereas standard models which neglect saturation fail especially for high currents. To consider the nonlinearities of the system directly in the control design, a nonlinear model predictive control scheme is used for controlling the magnet’s unstable dynamics. With focus on high-speed application, the nonlinear model predictive control algorithm is adapted to incorporate estimates of the guideway irregularities which are obtained from magnet control units located further to the front of the vehicle. The proposed control scheme offers much more flexibility in designing the controller focusing on ride comfort and gap variation, as shown by means of simulations. For these simulations guideway irregularities are applied which are derived from measurements. The necessary computations for the control law are performed in real-time.