Multivariable Identification of Active Magnetic Bearing Systems

Magnetic bearing systems are unstable MIM01 plants. In addition, if the levitated rotors are flexible, there may be poorly damped resonances of the levitated rotors. Advanced control is a crucial issue. System identification is therefore an important prerequisite for fast and reliable commissioning. Reported algorithms have difficulties to estimate the real unstable poles in magnetic bearing systems. Therefore a novel algorithm for multi variable identification has been developed. One of the key ideas is to identify the system poles from the determinant of the measured frequency response function matrix. The algorithm provides a state-space model of pre-defined order and structure, suited for controller design and verification. Experimental results with measured data from a magnetic bearing system with flexible rotor are included.