Premium
Identification and Self Tuning Control for Active Magnetic Bearing Systems: Levitation of Unknown Rotors
Author(s) -
Lösch F.,
Bühler Ph.
Publication year - 2002
Publication title -
pamm
Language(s) - English
Resource type - Journals
ISSN - 1617-7061
DOI - 10.1002/1617-7061(200203)1:1<242::aid-pamm242>3.0.co;2-n
Subject(s) - magnetic bearing , control theory (sociology) , rotor (electric) , gyroscope , controller (irrigation) , magnetic levitation , bearing (navigation) , helicopter rotor , engineering , identification (biology) , control engineering , system identification , rigid body , rigid rotor , rotordynamics , levitation , stiffness , control system , computer science , magnet , control (management) , mechanical engineering , aerospace engineering , physics , artificial intelligence , agronomy , botany , software engineering , structural engineering , classical mechanics , data modeling , biology , electrical engineering
The present paper deals with the problem of levitating rotors with unknown characteristics by means of active magnetic bearings whose properties are known. This problem is of interest in a technical setting to shorten the development time of AMB systems, in particular for controller design. Theoretical interest arises from the fact that several issues in the area of identification and self tuning control are addressed for an unstable system. Our aim is to identify the flexible rotor including gyroscopic effects and to automatically design a robustly stabilizing controller for this system that can be used for running the system under regular operating conditions. To this end, a rigid body model of the rotor is identified based on measured step responses from the plant. Then, the bearings are adjusted to have very low stiffness, and a controller with steep roll‐off is designed in order to avoid excitation of the unknown flexible modes of the system. Once the rotor is floating, the identification algorithm from [1] is applied to obtain information on the flexible modes of the system. Based on this extended model, a robust controller allowing for slow rotation of the rotor is designed. With the rotor rotating at a moderate speed, the frequency response functions are measured, and based on these measurements, the gyroscopic matrixof the system is identified, completing the system model and allowing for design of the desired controller for normal operation. The present contribution focusses on identification of the rigid body model of the flexible rotor.