On the Assessment of Slow Voltage Variations in Electric Distribution Networks using K-Means Clustering Algorithm

Gheorghe Grigoras, Bogdan-Constantin Neagu, Alexandra Adăscăliței

Abstract

The paper presents a clustering based analysis for the assessment of slow voltage variations from the electric distribution networks. The analysis uses the K-means clustering algorithm for classifying the electric substations (Medium Voltage/Low Voltage – MV/LV) in patterns characterized by linguistic terms from the point of view of slow voltage variations. The patterns were obtained using as input data in the clustering process the voltage quality indices (average deviation, average square deviation, and voltage irregularity degree). The analysis was made on a large supply zone with 722 electric distribution substations divided in three areas (remote, close, and suburb). The results confirmed the robustness and efficiency of proposed approach in the analysis of databases with a high amount of information insured by measurement devices.

Keywords

voltage quality,slow voltage variation, electric distribution networks, K-means clustering

Download

IMEKO-TC4-2017-026.pdf

DOI

-

IMEKO TC

TC4 - Measurement of Electrical Quantities

Back to the proceedings