Diagnosis of Disc Space Variation Fault Degree of Transformer Winding Based on K-Nearest Neighbor Algorithm

Winding is one of the most important components in power transformers. Ensuring health state winding great importance to stable operation system. To efficiently and accurately diagnose disc space variation (DSV) fault degree transformer winding, this paper presents a diagnostic method based on K-Nearest Neighbor (KNN) algorithm frequency response analysis (FRA) method. First, laboratory model used, DSV faults with four different degrees are achieved by changing discs winding. Then, series FRA tests conducted obtain results set up dataset. Second, ten numerical indices utilized features curves faulted Third, 10-fold cross-validation employed determine optimal k-value KNN. In addition, improve accuracy KNN model, comparative made between under distance functions. After getting appropriate metric k-value, classification constructed it used classify faults. The identification rate proposed 98.30%. Finally, performance presented comparing support vector machine (SVM), SVM optimized particle swarm optimization (PSO-SVM) method, random forest (RF). show that diagnosis highest can be