摘要
提出了基于AR模型和支持向量机的转子系统故障诊断方法.该方法对转子系统的振动信号建立AR模型,以AR模型主要的自回归参数和残差的方差作为特征向量,然后建立支持向量机分类器,进而判断转子系统的工作状态和故障类型.实验结果分析表明,该方法能有效地应用于转子系统的故障诊断.并通过支持向量机与BP神经网络的性能比较,说明了支持向量机的优点.
A fault diagnosis approach for rotor systems based on AR model and support vector machine is proposed. Firstly, the AR model of the vibration signal of a rotor system is established. The main auto-regressive parameters and the variances of remnant are regarded as the feature vectors. Then, the support vector machines used as fault classifiers are established to identify the condition and fault pattern of a rotor system. Practical examples demonstrate that the approach based on AR model and support vector machine can be applied to the rotor system fault diagnosis effectively. In comparison, the performances of support vector machine are more excellent than BP neural network.
出处
《系统工程理论与实践》
EI
CSCD
北大核心
2007年第5期152-157,共6页
Systems Engineering-Theory & Practice
基金
国家自然科学基金(50275050)
湖南省自然科学基金(05JJ40079)
中国博士后科学基金
关键词
AR模型
支持向量机
故障诊断
转子系统
AR model
support vector machine
fault diagnosis
rotor systems
作者简介
于德介（1957-），男，湖南大学教授，博士生导师．
