基于EEMD与RBF神经网络的齿轮故障诊断  

Gear fault diagnosis based on ensemble empirical mode decomposition and radial basis function neural network

作　　者：柴保明[1] 张江涛[1] 史朋波 张娴[1] 

机构地区：[1]河北工程大学,河北邯郸056038

出　　处：《煤炭工程》2017年第11期145-147,151,共4页Coal Engineering

摘　　要：文章在实验研究的基础上提出了EEMD与RBF网络相结合的齿轮故障诊断方法。重点讲述了EEMD的故障特征提取研究的方法和基本原理,简述径向基函数神经网络在故障诊断领域当中的应用。在故障模拟转子试验台上进行实验研究,采集足够的样本信号数据,利用集合经验模式分解原理来进行信号的特征提取,进而进行RBF网络的训练和故障识别,其较高的故障识别率验证了所提出的齿轮故障诊断模型的识别可靠性,同时该诊断模型具备着良好的应用前景。In this paper, a gear fault diagnosis method based on ensemble empirical mode decomposition and RBF neural network is presented. The fault feature extraction method of ensemble empirical mode decomposition is analyzed, and the application of RBF neural network in fauh diagnosis is introduced. The research experiment is carried out in the fault simulation test- bed, and the gear vibration diagnosis method based on EEMD and RBF neural network is validated by collecting the vibration signals of the gear. The experimental results show that the method is effective and has a broad application prospects.

关 键 词：齿轮 故障诊断 集合经验模式分解 径向基神经网络 

分 类 号：TH132.41[机械工程—机械制造及自动化]