User-Transformer Connectivity Relationship Identification Based on Shapelet Feature Extraction Method
Qingzhong Ni, Hui Jiang, Qirui Wang · 2023
In response to the issue of inaccurate records of topological information in low-voltage distribution networks, this paper proposes an voltage feature extraction method based on shapelet, coupled with the k-means clustering to identify user-transformer relationships automatically. First, the principal component analysis (PCA) algorithm is applied to reduce the dimensionality of the users’ voltage subsequences. Then, the most representative shapelets are selected according to distance assessment method, and the shapelet transform is performed to obtain voltage local features. Finally, clustering is employed to ascertain the connectivity relationship between users and transformers. The efficacy of the proposed method is validated by a practical case study in this paper.