论文部分内容阅读
由于铁路轴承的易损性和故障信号提取的复杂性,文中应用了多方法融合的诊断法对铁路轴承进行故障诊断,并对基于多方法融合的BP和RBF两种网络进行了比较。首先,对采集到得信号进行FIR降噪,再对降噪后的信号进行小波包分解,构造特征向量,以此为故障样本对BP和RBF网络进行训练,实现智能化故障诊断,实验结果表明文中提出的方法能很好地诊断出轴承故障类型,但多方法融合的RBF的泛化能力优于BP网络,同时,在训练时间上,RBF网络也要优于BP网络,这为机械故障诊断提供理论依据。
Due to the vulnerability of railway bearings and the complexity of fault signal extraction, a multi-method fusion diagnosis method is applied to fault diagnosis of railway bearings. BP network and RBF network based on multi-method fusion are compared. First, the FIR signal is denoised and the signal is decomposed by wavelet packet to construct the eigenvector. Then BP and RBF networks are trained as fault samples to realize intelligent fault diagnosis. The experimental results show that The proposed method can well diagnose the types of bearing faults, but the RBF with multi-method fusion is superior to BP network in generalization. At the same time, RBF network is superior to BP network in training time, Provide a theoretical basis.