Model Predictive Control for Unstable Processes -Application on Magnetic Suspension

Model predictive control is slowly but surly finding its way into industrial applications. While there are a lot of model predictive control algorithms for stable processes, the model predictive control algorithms for unstable processes are not so common. In this paper a design of predictive functional controller is presented that copes with stabilizing an unstable process. The design procedure is demonstrated on currently popular problem of magnet control. In the presented example the controller is designed to control a small scale magnetic suspension device where the magnet is used for positioning the metal rod to the desired distance.