New LMI based gain-scheduling control for recovering contact-free operation of a magnetically levitated rotor

© 2017 Elsevier Ltd A new approach for the recovery of contact-free levitation of a rotor supported by active magnetic bearings (AMB) is assessed through control strategy design, system modelling and experimental verification. The rotor is considered to make contact with a touchdown bearing (TDB), w...

全面介紹

Saved in:
書目詳細資料
Main Authors: M. Wang, M. O.T. Cole, P. S. Keogh
格式: 雜誌
出版: 2018
主題:
在線閱讀:https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85019062010&origin=inward
http://cmuir.cmu.ac.th/jspui/handle/6653943832/46641
標簽: 添加標簽
沒有標簽, 成為第一個標記此記錄!
實物特徵
總結:© 2017 Elsevier Ltd A new approach for the recovery of contact-free levitation of a rotor supported by active magnetic bearings (AMB) is assessed through control strategy design, system modelling and experimental verification. The rotor is considered to make contact with a touchdown bearing (TDB), which may lead to entrapment in a bi-stable nonlinear response. A linear matrix inequality (LMI) based gain-scheduling H ∞ control technique is introduced to recover the rotor to a contact-free state. The controller formulation involves a time-varying effective stiffness parameter, which can be evaluated in terms of forces transmitted through the TDB. Rather than measuring these forces directly, an observer is introduced with a model of the base structure to transform base acceleration signals using polytopic coordinates for controller adjustment. Force transmission to the supporting base structure will occur either through an AMB alone without contact, or through the AMB and TDB with contact and this must be accounted for in the observer design. The controller is verified experimentally in terms of (a) non-contact robust stability and vibration suppression performance; (b) control action for contact-free recovery at typical running speeds with various unbalance and TDB misalignment conditions; and (c) coast-down experimental tests. The results demonstrate the effectiveness of the AMB control action whenever it operates within its dynamic load capacity.