Integrated MIMO fault detection and disturbance observer-based control. (February 2021)

• Record Type: Journal Article
• Authors: Rouhani, Shahin; Tsao, Tsu-Chin; Speyer, Jason L.
• Is Part Of: Mechatronics. Volume 73(2021)
• Publication Date: 2021-02
• Subjects: Disturbance estimation -- State estimation -- Fault detection -- Game theory -- Optimization -- Observers -- MIMO systems -- Active magnetic bearing -- High speed machining; Computer integrated manufacturing systems -- Periodicals; Flexible manufacturing systems -- Periodicals; Mechatronics -- Periodicals; Productique -- Périodiques; Fabrication, Systèmes flexibles de -- Périodiques; Mécatronique -- Périodiques; Computer integrated manufacturing systems; Flexible manufacturing systems; Mechatronics; Periodicals; 629.89
• DOI: 10.1016/j.mechatronics.2020.102482
• Languages: English
• ISSNs: 0957-4158

Abstract

Abstract: This paper presents a new method of disturbance observer-based control (DOBC) for multi-input-multi-output (MIMO) highly-coupled unstable systems. In contrast to the current input–output approach for stable single-input-single-output (SISO) systems, the Youla parameterization of stabilizing controllers by full order state observer (FOSO) feedback control is shown more appropriate for general MIMO systems, while retaining the intuitive aspects of DOBC design. We propose a general MIMO DOBC expanded from this single FOSO stabilizing control, where a parallel number of FOSOs for fault detection, state estimation, and disturbance observer are integrated to achieve the relevant operational requirements and performance. Within this integrated control system, we propose a MIMO disturbance observer design method by a game-theoretic detection filter (GTDF) design. The DOBC design features GTDF disturbance decoupling followed by H-infinity model matching to establish the desired bandwidth for each channel of the decoupled disturbance observer. The proposed DOBC is applied to an open-loop unstable MIMO Active Magnetic Bearing Spindle (AMBS). Experimental results are presented to demonstrate the design method and control performance.