基于RBF人工神经网络的风机故障诊断  

Fault Diagnosis of Fan Based on RBF Neural Network

作　　者：李铁军[1] 朱成实[1] 叶龙[1] 赵新君[2] 王丹[1] 王学平[1] 

机构地区：[1]沈阳化工学院机械工程学院,沈阳市110142 [2]沈阳鼓风机(集团)有限公司

出　　处：《风机技术》2007年第5期71-73,共3页Chinese Journal of Turbomachinery

摘　　要：针对风机常见故障征兆与故障类型之间的非线性映射关系,结合专家知识建立了风机系统故障知识库,提出了基于径向基函数(RBF)人工神经网络的风机故障诊断方法。结果表明:该方法能克服诊断过程中容易陷入局部极小的缺点,并能满足故障诊断的快速性和准确性要求。According to the nonlinear mapping relationship between fault symptom and type of fan,the repository of fan faults system is established combined with experts' knowledge.The fault diagnosis method of fan based on RBF neural network is presented.The result shows that the method can overcome the limitation of local infinitesimal,and can meet the requirements for speediness and accurateness during fault diagnosis process.

关 键 词：RBF人工神经网络 故障诊断 风机 

分 类 号：TP11[自动化与计算机技术—控制理论与控制工程]