关键词: 无轴承永磁同步电机；神经网络；逆系统；解耦控制；径向悬浮力

Abstract:

The bearingless permanent magnet-type synchronous motor is a strong-coupled complicated nonlinear system.The linearization and decoupling control is a key to stable operation and practicability for this motor.Based on artificial neural networks,an inverse system decoupling control method for bearingless permanent magnet-type synchronous motors was proposed.This method combined approximation ability to nonlinear system,adaptability to parameter variations of artificial neural networks and decoupling characteristics of inverse system.Cascading the ANN inverse which consists of a static ANN and five integrators with the motor,the system was decoupled into two independent second-order linear subsystems and a first-order one,which were two displacement subsystems and a rotor speed one,so as to be easy to design the closed-loop linear regulator to control each of the subsystems.Simulation and experimental results show that strong robustness,good static and dynamic decoupling performance can be achieved by using the proposed method.

Key words: bearingless permanent magnettype synchronous motor, artificial neural networks;inverse system;decoupling control;radial levitation force