Volume 59

Smart City-Oriented Phase Identification for Unidirectional Users in Low-Voltage Distribution Networks Yucan Zhang, Yulin Chen2*, Donglian Qi1*, Xing Huang, Tianjian Zhang

Abstract

In smart city low-voltage distribution networks, the proliferation of residential microgrids, EV charging stations, and distributed generation sources induces significant three-phase imbalances in load, current, and voltage. These imbalances degrade power quality, reduce equipment utilization, and compromise system stability. Current phase identification methodologies rely on manual on-site inspections by utility personnel, entailing high labor expenditures, prolonged downtime, and operational inefficiencies. To this end, we propose a Smart City-Oriented Phase Identification method leveraging customer-side voltage data analysis. First, Principal Component Analysis (PCA) is leveraged to reduce dimensionality and extract salient features from voltage time-series data. Subsequently, Density-Based Spatial Clustering of Applications with Noise (DBSCAN) is innovatively integrated to cluster the PCA-derived feature vectors, enabling accurate inference of user-phase affiliations. Field validation conducted at a secondary substation of a local power supply bureau demonstrates real-time computational efficiency and achieves phase identification accuracy exceeding 90%.

Keywords low-Voltage distribution networks, phase identification, Principal Component Analysis (PCA), Density-Based Spatial Clustering of Applications with Noise (DBSCAN), Smart City Power Systems