基于KLLDA和ELM的新型模拟电路故障诊断方法

NOVEL ANALOG CIRCUITS FAULT DIAGNOSIS METHOD BASED ON KLLDA AND ELM

模拟电路是工业设备中最重要的元器件,其故障可能造成重大的人员伤亡,甚至造成巨大的经济损失。针对这一问题,提出一种基于核局部线性判别分析(Kernel Local Linear Discriminant Analysis,KLLDA)的故障诊断方案。利用小波分析和统计分析对原始信号进行预处理,得到原始特征集;利用KLLDA方法进行降维,并与核主成分分析(Kernel Principal Component Analysis,KPCA)和核线性判别分析(Kernel Linear Discriminant Analysis,KLDA)方法进行比较;采用极限学习机(Extreme Learning Machine,ELM)对测试电路的故障进行定位。对两个故障诊断案例的实验结果表明了该方法的有效性,并表明KLLDA在降维方面总体上优于KPCA和KLDA。

Analog circuit is the most important components in industrial equipment,and its failure may result in heavy causalities and even enormous financial losses.To cope with this issue,this paper proposes a fault diagnosis scheme based on kernel local linear discriminant analysis(KLLDA).Wavelet analysis and statistical analysis were used to preprocess the original signal to get the original feature set.Then,KLLDA method was utilized to reduce dimensionality,and compared with kernel principal component analysis(KPCA)and kernel linear discriminant analysis(KLDA)method.At last,we adopted the extreme learning machine(ELM)to locate the fault of the test circuit.The experimental results of two fault diagnosis cases show the effectiveness of the method,and KLLDA is superior to KPCA and KLDA in terms of dimensionality reduction as a whole.