Robust observer-based optimal linear quadratic tracker for five-degree-of-freedom sampled-data active magnetic bearing system. Issue 6 (26th April 2018)

• Record Type: Journal Article
• Authors: Tsai, Jason Sheng Hong; Su, Te Jen; Cheng, Jui-Chuan; Lin, Yun-You; Giap, Van-Nam; Guo, Shu Mei; Shieh, Leang San
• Is Part Of: International journal of systems science. Volume 49:Issue 6(2018)
• Page Start: 1273
• Page End: 1299
• Publication Date: 2018-04-26
• Subjects: Active magnetic bearing system -- observer/Kalman filter identification method -- equivalent input disturbance -- generalised linear quadratic digital tracker; System analysis -- Periodicals; 003.3
• DOI: 10.1080/00207721.2018.1443231
• Languages: English
• ISSNs: 0020-7721

Abstract

ABSTRACT: This paper presents three observer/Kalman filter identification (OKID) approaches and develops a robust observer-based optimal linear quadratic digital tracker (LQDT) for the five-degree-of-freedom (five-DOF) sampled-data active magnetic bearing (AMB) system with various disturbances. The more detailed objectives are: (i) to construct both an equivalent linear time-invariant discrete-time model and its state estimator via the proposed OKID approaches for the AMB system, which might be an unknown nonlinear time-varying unstable system with both a specified rotation speed and a sampling rate; (ii) to provide an adaptive disturbance estimation scheme, which establishes an equivalent input disturbance (EID) estimator for the AMB system with unexpected disturbances; and (iii) to develop a robust observer-based optimal LQDT for the sampled-data AMB system with both a pre-specified time-varying speed and unexpected disturbances. The developed LQDT is able to recover the displacement of the rotor to the pre-specified trajectory position whenever it deviates from such trajectory.