基于广义回归神经网络的采煤机故障诊断的研究  

Application of GRNN Network Used in Shearer Malfunction Fault Diagnosis

作　　者：尹丽娜[1] 汪琦[2] 樊后世[1] 梁鸿雁[1] 张会[1] 刘士阁[1] 赵世来[1] 

机构地区：[1]三一重型装备有限公司,沈阳110027 [2]沈阳施道克电力设备有限公司,沈阳110141

出　　处：《煤矿机械》2011年第5期244-245,共2页Coal Mine Machinery

摘　　要：针对采煤机故障征兆和故障的非线性对应关系,采用广义回归神经网络作为故障诊断的智能分类器。输入层为采煤机的故障特征参数,中间层为径向基神经元,感知待诊断故障向量与训练样本的相似度,输出层为故障模式分类。分析了广义回归神经网络的优越性和结构特征,建立了不同光滑因子和训练样本数目的采煤机故障诊断模型,并在MATLAB进行了仿真。Owing to the complicated relationship between the faults and the corresponding symptoms of shearer malfunction,this paper uses GRNN networks to diagnose fault.Fault character parameters are used as input layer cells,radius base cell is used as middle layer cells to detect comparability of undiagnosed vector and trained vectors,fault model is used as output layer cells.GRNN network＇s advantage and structure is analyzed,shearer malfunction fault diagnosis＇s model is built according to different smooth coefficient and samples.

关 键 词：采煤机 GRNN神经网络 故障诊断 

分 类 号：TD421.6[矿业工程—矿山机电] TP183[自动化与计算机技术—控制理论与控制工程]