基于支持向量机的磁力轴承系统辨识研究  

Active Magnetic Bearing System Identification Based on Support Vector Machine

作　　者：苏义鑫[1] 张丽[1] 蓝天[2] 

机构地区：[1]武汉理工大学自动化学院,湖北武汉430070 [2]华中科技大学电气与电子工程学院,湖北武汉430074

出　　处：《武汉理工大学学报（信息与管理工程版）》2012年第4期399-402,共4页Journal of Wuhan University of Technology：Information & Management Engineering

基　　金：湖北省自然科学基金资助项目(2009CDB082)

摘　　要：针对磁力轴承控制系统的设计分析,提出了一种基于支持向量机(SVM)的磁力轴承系统辨识方法。首先通过闭环控制使转子稳定悬浮,然后在控制器的输出信号中加入扰动信号,以使系统被充分地激励。在该闭环控制系统的基础上,对控制输入数据和输出数据进行采样,然后用SVM算法对磁力轴承系统进行辨识分析。该辨识系统的输入为控制电流,输出为转子位移。将该方法与BP神经网络进行比较,仿真结果表明,SVM用于磁力轴承系统辨识具有良好的辨识效果,辨识精度高,且训练速度快。A method of the active magnetic bearing system identification based on support vector machine （SVM） was proposed for meeting the demands of the system＇s analysis and control. For a stably suspended rotor,a random disturbance was added to the output of the closed loop controller in order to sufficiently excite the active magnetic bearing. In the model, the input parameter and the output parameter of the active magnetic bearing system were sampled, then the system was identified with SVM method. The control current was defined as the input parameter; the rotor＇s displacement was defined as output parameter. At the same time, the SVM method was compared with the BP method. Simulation results show that, SVM for the identification of magnetic bearing system has a good recognition results, which provides a reference method for the controller of the active mamaetic bearing.

关 键 词：支持向量机 磁力轴承 非线性系统 系统辨识 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]