基于神经元网络故障诊断主特征量提取的应用研究  

Application of Picking up Important Characteristic Values in Fault Diagnosis Based on Neutron Network

作　　者：郑发泰[1] 

机构地区：[1]浙江工商职业技术学院,浙江宁波315012

出　　处：《煤矿机械》2007年第6期188-190,共3页Coal Mine Machinery

摘　　要：分析了基于神经元网络故障诊断主特征量提取的方法,论述了几何距离函数和信息距离函数算法的构成及在神经元网络分类器中的应用。通过实验表明,在运用神经元进行故障诊断中采用主特征量提取法对特征量进行信息压缩是行之有效的。Analyzes how to pick up the important characteristic values in fault diagnosis based on neutron network,also describes how the algorithm of geometric distance function and the information distance function is configurated and its application in the neutron element network segregator.The test shows that it is an effective method by picking up the important characteristic values to compact the information in neutron element fault diagnosis.

关 键 词：神经网络 故障诊断 主特征量 液压系统 

分 类 号：TP306[自动化与计算机技术—计算机系统结构]