LS-SVM Inverse System Decoupling Control Strategy of a Bearingless Induction Motor Considering Stator Current Dynamics

Wenshao Bu,Ziyuan Li

doi:10.1109/access.2019.2939258

Abstract

The bearingless induction motor (BL-IM) is a multivariable, nonlinear and strongly coupled object, in order to achieve its dynamic decoupling control with high performance, on the basis of the least square support vector machine(LS-SVM) principle, a novel LS-SVM inverse system decoupling control strategy is proposed. Firstly, under the conditions of taking the stator current dynamics of torque windings into account, the state equations of the BL-IM system are established. Secondly, based on the approximation and identification fitting ability of the LS-SVM to arbitrary nonlinear function, the inverse system mathematical model of the BL-IM system considering the stator current dynamics is trained and obtained. After that, according to the decoupling principle of inverse system, the BL-IM system is decoupled into four second-order pseudo-linear integral subsystems, include a motor speed subsystem, a rotor flux-linkage subsystem and two radial displacement component subsystems. At the end, the comprehensive simulation analysis of the LS-SVM inverse decoupling control system are carried out. From the simulation results, it can be known that the dynamic decoupling control among the motor speed, rotor flux-linkage and two radial displacement components can be realized, meanwhile after considering the dynamics of the stator current, the LS-SVM inverse decoupling control system of BL-IM has the characteristics of fast response and strong anti-interference ability.

Highlights

The AC motor supported by magnetic bearing has been widely used in the high speed drive field [1]–[4], but it still has some disadvantages, such as the higher magnetic suspension power consumption, the difficulty to over speed, etc. [5]–[8]
In order to control the bearingless induction motor (BL-IM) with high-performance, this paper proposes a LS-support vector machine (SVM) inverse decoupling control strategy for the BL-IM system
3) CONTROL SYSTEM SYNTHESIS After the off-line trained least squares support vector machine (LS-SVM) inverse system is combined with the original BL-IM system, the BL-IM system is decoupled into four pseudo-linear second-order integral subsystems, include a motor speed subsystem, a rotor fluxlinkage subsystem and two radial displacement component subsystems [22]

Summary

INTRODUCTION

The AC motor supported by magnetic bearing has been widely used in the high speed drive field [1]–[4], but it still has some disadvantages, such as the higher magnetic suspension power consumption, the difficulty to over speed, etc. [5]–[8]. C. STATE EQUATION OF A BL-IM SYSTEM Taking the current dynamics of the torque system into account, the input-, state- and output-variable of the BL-IM system are defined as follows:. The inverse system model that takes the current dynamics of torque system into account can be expressed as follow:. In this paper, in order to overcome the mentioned shortcomings of the analytic inverse system method, utilizing the stronger self-learning and generalization ability of the LS-SVM, the LS-SVM inverse model of a BL-IM system that takes the current dynamics of torque system into account would be identified. The LS-SVM topology that takes other variable as the output variable is similar with this, and no further details will be given here

LS-SVM INVERSE SYSTEM DECOUPLING CONTROL

SYSTEM SIMULATION VERIFICATION AND ANALYSIS

CONCLUSION