摘要
分析了小波分析、模糊聚类以及神经网络各自在信号处理方面的优势,以小波变换(或小波包变换)提取故障特征,对神经网络的输入和输出进行模糊化处理,最后用神经网络进行诊断,从而提出了一种有别于传统的、高效的机械故障诊断方法。
We analyze the advantages of wavelet analysis, fuzzy clustering and neural network in the signal processing respectively. The malfunction characteristic is abstracted with wavelet transform, and the entering and output of neural networks is carried out fuzzification, then neural networks is used to carry out a diagnose. Different from traditional one, an effective malfunction diagnostic method has been proposed based on wavelet fuzzy clustering and neural network.
出处
《轻工机械》
CAS
2008年第6期122-123,129,共3页
Light Industry Machinery
基金
陕西省教育厅产业化培育项目(07JC05)
咸阳市科技计划项目(XK0708-6)
陕西科技大学博士科研启动基金(BJ06-05)
关键词
小波分析
模糊聚类
神经网络
模式识别
故障诊断法
wavelet analysis
fuzzy clustering
neural network
pattern recognition
malfunction diagnose method
