数控机床高速无轴承异步电动机悬浮子系统RBFNN逆独立解耦控制

Independent RBFNN inverse decoupling control for levitation subsystem of bearingless induction motor of NC machines

为了实现对数控机床用高速无轴承异步电动机(Bearingless Induction Motor,BIM)动态解耦控制以实现降低悬浮电主轴抖动,提出了一种基于径向基函数神经网络(Radial Basis Function Neural Network,RBFNN)的悬浮子系统自适应独立控制方法。首先,基于RBFNN构建了气隙磁链观测器,因为RBFNN具有较强的自学习和自适应能力,所以辨识的气隙磁链较为精确;其次,基于Hamilton-Jacobi-Isaacs(HJI)原理设计RBFNN逆系统鲁棒控制器,应用基于HJI不等式的RBFNN辨识系统模型不确定和外界干扰,提高系统的稳定性,悬浮子系统动态独立解耦控制得以实现;最后,将磁链辨识器和逆系统鲁棒控制器组成双RBFNN悬浮子系统逆独立控制系统。仿真和实验结果表明,采用该控制方法 BIM系统能获得良好的动、静态性能。

To realize the dynamic decoupling control of high speed bearingless induction motor （BIM）of NC machines,a self-adaptive independent control method based on radial basis function neural network （RBFNN）was proposed here.Firstly,with this method,an air-gap flux observer was built to obtain a more accurate air-gap flux identifier due to stronger self-learning and self-adaptive ability of RBFNN.Furthermore,a RBFNN self-adaptive robust controller based on Hamilton-Jacobi-Isaacs （HJI ） was designed to realize the decoupling control of the levitation subsystem of BIM stably and reliably.Finally,the self-adaptive independent control system with double RBFNN was composed of the proposed air-gap flux observer and the self-adaptive robust controller.The simulation results showed that the proposed system has good dynamic and static performances;in addition,this proposed method not only realizes the decoupling control of torque and radial suspension forces,but also realizes that of radial suspension forces in both two degrees of freedom of the system.