摘要
滚动轴承是旋转机械中应用普扁而又易损的元件之一，其故障在机械故障中占有很大的比例．因此，轴承故障诊断、特别是早期诊断很受重视．本文将神经网络应用于轴承早期故障诊断，简要说明了ＢＰ神经网络的基本原理、算法及特点，介绍了实验数据的分析过程和参数选择原则．实验结果表明，选择适当的网络结构进行训练、学习和检验，可以把良好轴承、内环缺陷轴承、外可缺陷轴承、滚子缺陷轴承及具有三种综合缺陷的轴承区分开来，并能初步估计出缺陷的大小．
Rolling element bearing is one of the most commonly used and easily damaged component in the machines. Many of the machinery the machinery failures are attribuble to faulty rolliing element bearings,so the fault disgaosis of rolling element bearing, especially early fault diagnosis important. The paper presents an application of the Artitfical Neural Networks(ANN) in the early fault diagnosis forrolling element bearings. The backpropagation of the ANN, the procedure to process data and the thinking to select the parameters are introduce bristly. The ANN was trained and testified with those data. It's proved that the ANN method is efficient to be used in the early fault diagnosis of rolling element bearings. A neural network with proper size can distinguish the following defective bearing conditions from each other no fault, localized sciling fault of inner outer race or one ball and the stwhesis of the cave mentional three kinds of faults.
出处
《机械科学与技术》
CSCD
北大核心
1996年第3期464-467,共4页
Mechanical Science and Technology for Aerospace Engineering
基金
航空科学基金
关键词
故障诊断
滚动轴承
神经网络
Fault diagnosis Rolling element bearing Neural network
