摘要
提出一种基于小波熵和自适应神经模糊推理系统(ANFIS)的滚动轴承故障诊断方法。对原始振动信号进行小波包分解,提取小波熵特征,然后用ANFIS作为分类器进行故障模式识别。比较人工神经网络(ANN)与ANFIS的故障诊断效果。结果表明,该算法诊断准确率高于神经网络。
A fault diagnosis method of rolling bearings based on wavelet entropy and adaptive neuro-fuzzy inference system（ANFIS）was presented.The vibration signal was decomposed using wavelet packet,and wavelet entropy was calculated and selected as feature.ANFIS model was used as classifier to identify different faults.For comparison purposes,artifical neural network method was investigated.The results indicate that the rate of accuracy of fault diagnosis of ANFIS model is higher than that of artifical neural network.
出处
《机床与液压》
北大核心
2010年第11期137-139,共3页
Machine Tool & Hydraulics
基金
山东省自然科学基金资助项目(Y2005F12)
关键词
故障诊断
自适应神经模糊推理系统
小波熵
神经网络
Fault diagnosis
Adaptive neuro-fuzzy inference system（ANFIS）
Wavelet entropy
Neural networks
作者简介
隋文涛（1977-），男，博士研究生，讲师，研究方向为故障诊断、信号处理。电话：13969324462，E—mail：suiwt@163．com
