发电机螺栓松动故障特征检测方法及试验研究

提出一种通过提取时域指标特征和运用主成分分析法(PCA)诊断螺栓松动故障的方法,将原始数据进行经验模式分解(EMD)后计算相应IMF的5个无量纲因子;利用主成分分析法(PCA)对数据向量进行降维和残基空间投影处理,计算数据样本的预测误差,在智能供水系统试验台架上开展螺栓松动故障诊断试验,试验结果表明,所建立PCA模型能够较为有效判别螺栓松动故障。

Research on Detection Method and Test of Generator’s Bolt Loose Faults

Based on the technique of time-domain index feature extraction and principal component analysis(PCA),a method for fault diagnosis of loose bolt is proposed.Five dimensionless factors corresponding to IMF are calculated after the empirical mode decomposition(EMD)of the original data.Then,the PCA is applied to the data vectors for dimension reduction and residual space projection.And the prediction error of the data sample is calculated.At last,fault diagnosis test of bolt loosening is carried out on the test bench of an intelligent water supply system.The results show that the proposed PCA model is effective in judging the bolt loosening faults.