Model Predictive Control for Active Suspension of Electrodynamic Levitation System with Augmented Kalman Filter

Suspension control stands as a fundamental technique in magnetic levitation (Maglev) trains, where ensuring the comfort of passengers is paramount. The increasing demand for fast, safe, and efficient transportation systems in recent years has made the design of high-velocity trains, such as Maglevs, inevitable. This project, first addresses the issue of magnetic pad oscillations. Further, a state-of-the-art Model Predictive Controller (MPC) is introduced, which relies on prior knowledge of the system's states for prediction. Additionally, an augmented Kalman filter (AKF) is employed to estimate the desired states for control. A novel approach is employed to solve the issue regarding the drift in the system state estimation. The performance of the latter is then compared to experimental results obtained from a test bench. Ultimately, it is demonstrated that MPC outperforms passive damping and active suspension with a static gain controller (LQR).