基于离散BAM网络的汽轮发电机组振动故障诊断的应用研究  

Fault Diagnosis of Turbogenerator Units Vibration Based on Discrete BAM Neural Network

作　　者：彭希[1] 刘涤尘[1] 

机构地区：[1]武汉大学电气工程学院,湖北武汉430072

出　　处：《大电机技术》2005年第3期13-16,共4页Large Electric Machine and Hydraulic Turbine

摘　　要：本文针对常规频谱诊断方法的不足,详细论述了离散的BAM网络及其特性。进而讨论了汽轮发电机组常见典型振动故障的变化特征及其数字化描述方法,建立了基于BAM算法的神经网络的汽轮发电机组振动故障诊断模型,对实际机组故障进行了诊断,效果良好。通过分析,提出BAM网络在故障诊断领域中有广阔的应用前景。Based on some shortages of conventional diagnostic methods, discrete BAM neural network is particularly discussed in this paper. And various vibration characteristics of typical faults on turbogenerator units are discussed. Then,the neural network model for diagnosing vibration fault of turbogenerator sets based on BAM algorithm is established and used for practical fault and good results are gained. As a result,broad application prospect of BAM neural network method in the field of fault diagnosis can be envisaged.

关 键 词：汽轮发电机组 BAM网络 振动故障诊断 应用 离散 故障诊断模型 诊断方法 描述方法 变化特征 机组振动 神经网络 数字化 

分 类 号：TM311[电气工程—电机] TP18[自动化与计算机技术—控制理论与控制工程]