•  
  •  
 

Keywords

distributed power supply; affinity propagation clustering algorithm; direct power supply; correlation analysis

Abstract

The randomness and intermittency of the output of distributed photovoltaic power sources pose significant challenges to the stable operation and power quality of distribution networks. In this paper, a method based on affinity propagation clustering algorithm (APCA) is proposed to effectively identify the direct power supply situations of photovoltaic power sources. Unlike traditional methods, photovoltaic power sources in the same station area are selected by this algorithm, which reduces the influence of factors such as climate and light intensity and improves the accuracy and operability of the algorithm. First, correlation coefficients are used to screen out photovoltaic power sources with high correlation. Secondly, cluster analysis is conducted on the three-phase grid-connected power curves using the AP clustering algorithm, which further identifies suspected direct power supply and determines possible direct power supply types. Finally, a case analysis is used to verify that the proposed method helps power supply companies better manage distributed photovoltaic power source systems and improves the fairness and transparency of the electricity market.

DOI

10.19781/j.issn.1673-9140.2026.02.018

First Page

201

Last Page

211

Download

Share

COinS