Nonlinear Modeling of a Magnetic Bearing System using SDP model and Linear Wavelet Parameterization

In this paper, we present the nonlinear modeling of a magnetic bearing system. The approach uses the SDP model to characterize the system´s nonlinearities; and an optimized model structure selection algorithm in which linear wavelet functional approximation and PRESS statistics are in use prior to the final parametric estimation. The experimental results reported in this paper demonstrate the approach in identifying a test-stand magnetic bearing system.