@Article{ee.2025.072828,
AUTHOR = {Jian Wang, Gongqiang Yang, Yufeng Sun, Gangui Yan, Jie Long},
TITLE = {Distribution Network Partitioning and Distributed Voltage Coordinated Optimization Method under High-Proportion Photovoltaic Penetration},
JOURNAL = {Energy Engineering},
VOLUME = {},
YEAR = {},
NUMBER = {},
PAGES = {{pages}},
URL = {http://www.techscience.com/energy/online/detail/25153},
ISSN = {1546-0118},
ABSTRACT = {Given that the power grid partitioning method relying mainly on line reactive power flow information sees frequent changes in partitioning results with reactive power flow fluctuations under high-proportion fixed-power-factor PV-connected distribution networks, and traditional distributed PV collaborative optimization fails to adapt due to such changes, a stable partitioning and distributed PV collaborative optimization method for this scenario is proposed. Firstly, the Gaussian mixture model (GMM) is used to characterize the characteristics of PV reactive power output, obtaining the typical curve of PV reactive power output. Secondly, the Monte Carlo Simulation (MCS) probabilistic power flow calculation is performed to obtain the node voltage distribution of the distribution network. Thirdly, based on the node voltage distribution, the Earth Mover’s Distance (EMD) is used to obtain the statistical distance between any two nodes, and this statistical distance is combined with the electrical distance defined by node voltage sensitivity to form a comprehensive electrical distance. Then, the affinity propagation clustering algorithm is applied, and considering the dynamic reactive power margin requirement, the reactive power/voltage partitioning result is obtained. Based on the reactive power partitioning result, a reactive power optimization model is established with the minimum active power loss of the system as the objective function. The optimization model is convexified using the LinDistFlow equation, and the Alternating Direction Multiplier Method (ADMM) is adopted to coordinate the reactive power output of PV inverters in each partition, achieving global optimal voltage control in the distribution network. Finally, the proposed method is verified using the IEEE 33-bus system. The application of this method reduces the system power loss by 35.94%. Compared with the traditional partitioning method, the partitioning variation rate under Scenario 1 is reduced by 54.17% and that under Scenario 2 is reduced by 70.85% when this method is adopted. This fully demonstrates that the partitioning results of the proposed method are stable, and the collaborative optimization method can improve the system voltage stability and reduce the system power loss.},
DOI = {10.32604/ee.2025.072828}
}