Application of Principal Components Analysis in Condenser Fault Diagnosis

Condenser faults are usually caused by a variety of factors. The symptoms of different faults are characterized by their similarities. The relevance of such similar symptoms produces difficulties of data analysis. The parameters that impact condenser faults are comprehensively optimized by using principal components analysis (PCA) in this paper so that the information of feature parameters can be fused to several principal components, and then the dimensions of standard fault patterns can be determined according to the contribution ratios of each principal component, it is proved that the method is valid by cluster analysis for the samples to be diagnosed