Research on representative day selection method based on grey relational and K-means clustering

With the rapid development of renewable energy technology, Distributed Generation (DG) has become an important component of the power system. The integration of distributed power sources not only improves the flexibility and reliability of power systems, but also brings new challenges to load forecasting, grid scheduling, and line loss calculation in power systems. In response to the problem of selecting representative days for theoretical line loss calculation, this paper considers the impact of distributed power generation on the power system and proposes a representative day selection method based on grey correlation and k-means clustering. Firstly, the grey correlation method is used to select the top ten indicators that are related to the daily line loss rate. Then, the K-means clustering analysis method is used to cluster the top ten indicators, divide the substation data into three categories, select the most effective clustering center, and calculate the distance between the substation data and the clustering center on a monthly basis. The closest day to the clustering center is used as the representative day of this month. Finally, experimental data from Zhuhai was selected to verify the effectiveness of the proposed method.