摘要
转子系统中的振动信号包含了很多状态信息,运行过程中故障特征的有效提取和识别对于转子系统早期故障诊断非常关键。针对转子系统故障信息的复杂性,提出将小波包分析和支持向量机相结合的转子系统早期故障诊断方法。该方法首先利用改进的小波包方法提取早期故障特征;然后将提取的特征向量输入基于支持向量机的分类器进行故障识别。实验分析结果表明,该方法在小样本情况下,能够有效识别转子系统的早期故障,具有很好的分类精度,而且能够实现旋转机械的多故障诊断。
The vibration signals of rotor in operation tion, and extraction and identification of fault fault diagnosis signals consist of plenty of information about its running condiin the process of speed change are necessary for the early of rotor system. Due to the complexity of fault diagnosis for rotor, a new method for early fault diagnosis on rotor system is proposed which combines the wavelet packet and support vector machine. Firstly, the improved wavelet packet is used to extract early fault feature signals. Then, the early fault feature vector is inputted to the classifier based on the support vector machine. The results show that the proposed method can identify the early fault of rotor system in small sample, classification precision is satisfactory, and the multi-faults diagnosis of rotor system is realized.
出处
《测控技术》
CSCD
北大核心
2014年第9期18-21,25,共5页
Measurement & Control Technology
基金
中国博士后科学基金资助项目(2014M552504)
关键词
故障诊断
小波包
支持向量机
粒子群算法
fault diagnosis
wavelet packet
support vector machine
particle swarm algorithm
