基于神经网络的车辆轴承故障诊断技术  

The Trouble Diagnostics Technology for Railway Car Bearings on the Basis of Nerve Network

作　　者：丁福焰[1] 张生玉[1] 邵军[1] 

机构地区：[1]铁道科学研究院机车车辆研究所,北京100081

出　　处：《铁道车辆》2006年第12期7-11,共5页Rolling Stock

摘　　要：简要介绍了车辆轴承轨边声学诊断技术和神经网络诊断方法,重点讨论了BP网络在车辆滚动轴承故障诊断中的应用,用大量的现场实测数据作为训练样本,对所建立的BP网络进行训练,将训练好的网络应用于轴承状态分类,具有很高的准确性,说明神经网络方法是车辆轴承故障诊断的一种有效解决方案。The Trackside acoustic diagnostic technology and the nerve network diagnostic method for railway car bearings are briefly described.The application of BP network in the trouble diagnostics for railway car bearings is mainly discussed.A lot of actually measured data in field are used as the training samples to make training for the BP network already founded,and the trained network is applied in the bearings status classification,which has high accuracy.Hence,it is shown that the nerve network me-thod is an effective solution scheme in the trouble diagnostics for railway car bearings.

关 键 词：铁道车辆 轴承 故障诊断 神经网络 

分 类 号：U279.323[机械工程—车辆工程]