Power Transformer Anomaly Detection Based on Adaptive Kernel Fuzzy C-Means Clustering and Kernel Principal Component Analysis

In this paper, a data-driven power transformer anomaly detection method is presented. For transformers with multiple working states and nonlinearity of data, adaptive kernel fuzzy C-means clustering (KFCM) algorithm is used to cluster sample data, each class corresponds to a working state. Then, the kernel principal component analysis (KPCA) is used to obtain projection matrices and anomaly detection limits of various classes. A method for online update of sample data is designed in this paper. Finally, the measured data of the power transformer is used for experiment, and the results are compared with the results obtained by using the conventional KPCA algorithm. The experimental results prove the correctness and effectiveness of the method in this paper.