风电机组齿轮箱故障分类方法研究  

Research on Fault Classification Method of Wind Turbine Gearbox

作　　者：李状[1] 马志勇[1] 姜锐[1] 柳亦兵[1] 

机构地区：[1]华北电力大学能源动力与机械工程学院,北京102206

出　　处：《机械设计与制造》2015年第2期177-180,共4页Machinery Design & Manufacture

基　　金：国家自然科学基金(51305135);中央高校基本科研业务费专项资金资助(2014XS15)

摘　　要：风电机组齿轮箱的运行工况复杂多变,很难获取大量的所有已知故障的样本数据,为了能够实现在无已知样本数据条件下的故障分类,提出了一种基于ART2神经网络和C-均值聚类算法的风电机组齿轮箱故障分类方法。首先利用ART2无监督神经网络实现样本数据的初步分类,再利用C-均值聚类算法对分类结果进行修正,克服了由于原始神经网络算法存在"硬竞争"导致分类精度下降的问题。分析结果表明提出的方法具有更高的分类准确度,能够对健康和不同故障类型的风电机组齿轮箱进行准确分类识别。The operating conditions of wind turbine gearbox are complex and changeable. It is difficult to acquire a large number of training samples for all known faults. For implementing fault classification without known samples, a method with ART2 neural network and C-mean clustering algorithm was propased for fault diagnosis of wind turbine gearbox. Firstly, ART2 unsupervised neural network was used for initial classification. Then C-means clustering algorithm was introduced to modify the classification of samples and the problem of low classification accuracy caused by the hard competition of ART2 was solved. Results show that the proposed method has higher classification accuracy and can classify and recognize the normal gearbox and that of different fault types of wind turbine.

关 键 词：风电机组 齿轮箱 ART2神经网络 C-均值聚类 无监督分类 故障诊断 

分 类 号：TH16[机械工程—机械制造及自动化] TH165.3