Magnetic Bearings and Non-Stationary Dynamics of Rotors

This contribution is an attempt to spur interest in the application of neural networks in actively supported rotors. After some introductory remarks non-stationary rotordynamics is reviewed in short, restricted to the jeffcott Rotor (section 2) . Section 3 than gives an overview of a simple neural network trained by the backpropagation procedure. The latter is then utilized in section 4 to avoid suspended stays of the non-stationary rotor. Section 5 outlines future improvements, realizations and other possible applications. A short summary concludes the contribution.