Decoupling control for bearingless synchronous reluctance motor based on support vector machine inverse. (2nd March 2015)

• Record Type: Journal Article
• Authors: Yang, Zebin; Sun, Xiaodong; Zhu, Huangqiu
• Is Part Of: International journal of modelling, identification and control. Volume 23:Number 1(2015)
• Page Start: 24
• Page End: 32
• Publication Date: 2015-03-02
• Subjects: bearingless synchronous reluctance motor -- BSRM -- mathematical model -- least squares support vector machine -- decoupling control; Engineering -- Methodology -- Periodicals; Science -- Methodology -- Periodicals; 001.42
• Languages: English
• ISSNs: 1746-6172

Abstract

A novel decoupling control method based on least squares support vector machine inverse (LS-SVM) is presented for a bearingless synchronous reluctance motor (BSRM) possessing the characteristics of multi-input-multi-output, nonlinearity, and strong coupling. Based on the analysis of operation principle, the dynamic mathematical models are built, which are verified to be invertible using interactor algorithm. The nonlinear inverse model is then obtained offline using LS-SVM for its ability of function fitting. The LS-SVM inverse is connected in series with original system to decouple the BSRM system to three single-input-single-output pseudo linear systems. To further compensate the modelling error to improve robust performance, the PID feedback controller is designed, which together with inverse feedforward controller gives a composite control strategy. The simulation and experimental results show that the presented LS-SVM inverse control strategy can realise the dynamic decoupling of bearingless synchronous reluctance motors. And the control system has fine dynamic and static performance.