摘要
风电机组传动链故障特征类型丰富,故障信息量大,提出将振动信号作为其故障特征信息的载体,通过小波包变换进行特征提取。应用RBF神经网络的非线性映射能力和自适应学习能力,以提高故障诊断的准确性,详细分析了进行两级神经网络识别的思想,应用第一级神经网络进行故障分类,应用第二级神经网络进行具体故障识别。给出了故障诊断试验,该试验充分说明应用小波包算法进行故障特征提取可以有效地实现故障信号的预处理,进行两级神经网络可以有效提高故障诊断效率。
Wind turbine driving chain with abundant fault feature and variable types,the vibration signal w as a carrier of fault features w hich can effectively reflect most of the fault information of the w ind turbine drive train.Wavelet packet transform w as adopted for feature extraction.With the application of nonlinear mapping and adaptive learning feature of neural netw orks to enhance its accuracy,the tw o-level neural netw ork recognition method w as proposed,w ith first level for fault classification and second level for fault diagnosis.The example show s that this method can be effectively applied to drive train of w ind turbine fault diagnosis w ith w avelet packet algorithm w ith fault feature extraction and tw o-level neural netw ork pattern recognition.
出处
《组合机床与自动化加工技术》
北大核心
2013年第10期95-97,共3页
Modular Machine Tool & Automatic Manufacturing Technique
基金
河南省科技厅科技攻关计划资助项目(122102210416
112102210339)
关键词
传动链
小波包
RBF神经网络
故障诊断
特征提取
drive train
wavelet packet
RBF neural networks
fault diagnosis
feature extraction