Eigenvalue analysis for a magnetic levitation mover using camera-derived time series via system model subjected to noisy excitation

Understanding and improving vibration suppression, as well as detecting coil failures in magnetic levitation systems, are important for industrial applications. This paper conducts an operational modal analysis of the closed-loop system, focusing on the frequency region below 100 Hz. The frequency response analysis methodology is developed using the displacement signal derived from the Vicon camera data in the maglev system. This study reveals the root locus and controllability associated with the levitated height and in-plane location of the maglev mover. The eigenvalue analysis of the maglev system was conducted by introducing a coupled PID model and a corresponding identification method using the frequency response spectrum data. The coupled PID model was introduced to describe the random vibration of the closed-loop maglev system at the static levitated state. The algorithm combines the frequency curve fitting approach and the analytical eigenanalysis-based initial value setting method for the coupled PID model. From the analysis of the polar plots and the root loci, the majority of the eigenmodes become unstable as the levitated height increases. In a part of the high-frequency modes, the eigenmodes were expected to become stable. Furthermore, the controllability was verified using the Gramian matrix. It was observed that state controllability is well achieved under high levitation conditions. However, this implies a trade-off is adopted between system stabilization and controllability.