摘要
文中将神经网络技术运用于滚动轴承故障诊断．着重讨论了传统故障诊断技术所面临的挑战及基于神经网络的故障诊断技术的优越性．传统的故障诊断专家系统在知识的获取及表达上存在着困难，并且当系统较大时，变得非常庞大，不适于在线控制；人工神经网络对信息分布式的存储及处理，具有明显的优点．以滚动轴承故障诊断为例，编制了基于神经网络技术的诊断程序，运行结果表明，该方法具有判断准确、容错性好、适于在线工作、便于推广的优点．
The paper the neural networks technology has been applied to the ball bearing failure diagnosis． Discuses the present situation of traditional failure diagnosis and the advantage of neural networks failure diagnosis． Traditional failure diagnosis expert system has trouble in getting and expressing knowledege and become very big when the system is big． It cannot work at line． Artificial neural networks has obvious advantage in storing and dealing with knowledege． In the end drawsup a diagnose program of neural networks failure diagnosis． This methode can judge accuratly work at line and is easy for diffusion．
出处
《南方冶金学院学报》
2001年第2期141-145,共5页
Journal of Southern Institute of Metallurgy
关键词
故障诊断
神经网络技术
反向传播算法
滚动轴承
mechanical failure diagnosis neural networks technology back propagation algorithm
