Abstract: In view of the low efficiency of traditional voltage quality abnormal identification methods, it is difficult to achieve comprehensive identification and monitoring. This paper builds a platform model for voltage quality diagnosis and analysis. The principal components of applications such as voltage monitoring and analysis of voltages and operation status analysis of end users in distribution network low-voltage station areas are obtained through principal component analysis. Clustering analysis screens out abnormal voltage data that meets the characteristics of abnormal voltage, and uses abnormal voltage model to determine abnormal data to generate abnormal voltage recognition results. Based on the J2EE technical framework, visual display is made in terms of end-user voltage monitoring analysis, distribution network low-voltage station operation status analysis, and line fault judgment analysis of distribution network.