沈煜, 李穆, 杨志淳, 杨帆. 考虑负荷分布特征的供电单元聚类优选方法[J]. 电网与清洁能源, 2023, 39(4): 54-62.
引用本文: 沈煜, 李穆, 杨志淳, 杨帆. 考虑负荷分布特征的供电单元聚类优选方法[J]. 电网与清洁能源, 2023, 39(4): 54-62.
SHEN Yu, LI Mu, YANG Zhichun, YANG Fan. A Clustering Optimal Division Method for Power Supply Units Considering Load Distribution Characteristics[J]. Power system and Clean Energy, 2023, 39(4): 54-62.
Citation: SHEN Yu, LI Mu, YANG Zhichun, YANG Fan. A Clustering Optimal Division Method for Power Supply Units Considering Load Distribution Characteristics[J]. Power system and Clean Energy, 2023, 39(4): 54-62.

考虑负荷分布特征的供电单元聚类优选方法

A Clustering Optimal Division Method for Power Supply Units Considering Load Distribution Characteristics

  • 摘要: 科学合理的供电单元划分是中压配电网接线模式优选的关键,传统的供电单元划分方法通常基于平均负荷密度和供电半径,难以处理海量的配电网精确负荷空间分布信息(例如经度、纬度、负荷等)。提出了一种考虑负荷分布的聚类供电单元优选方法。首先基于供电网格内的负荷的空间坐标以及负荷,构建精确负荷空间分布数据样本集合。定义负荷密集度指标量化供电单元内部负荷分布的紧密程度,定义单元离散度指标量化供电单元之间界限的清晰度,并构建单元优选目标函数,基于样本点典型日负荷曲线的最大功率,采用聚类算法进行遍历优选,得到供电单元的初始划分方案;基于技术指导性原则定义相应的约束条件,包括最大供电半径、负荷总量以及容载比,进一步考虑负荷的时序波动特性,对初始划分方案进行调整,使得满足约束条件的供电单元内部负荷分布尽量紧密而供电单元之间的界限尽量清晰。最后结合某供电网格实际负荷空间分布三维信息作为算例,分别采用K-means以及DPC聚类算法验证了所提方法有效有效考虑到负荷的实际空间分布,划分出的供电单元内部负荷分布紧密,供电单元之间界限清晰。

     

    Abstract: Scientific division of power supply units is the key to the selection of the wiring mode of medium-voltage distribution networks. The traditional division of power supply units is usually based on the average load density and power supply radius,and it is difficult to deal with the massive amount of precise load spatial distribution information(such as longitude,latitude,and load size,etc.). In this paper,a clustering division method for power supply units considering load distribution is proposed. First,based on the spatial coordinates and load size of the load in the power supply grid,a sample set of accurate load spatial distribution data is constructed. Second,the load intensity index is defined to quantify the tightness of the load distribution within the power supply unit,and the unit dispersion index is defined to quantify the clarity of the boundaries between the power supply units,and the objective function is constructed,and the clustering algorithm is used to traverse and optimize,and obtain the initial power supply unit division scheme. Further considering the technical guidelines,the constraint conditions are defined,including the maximum power supply radius,capacity-load ratio and total load. The initial division scheme which meets the technical guidance is adjusted to make the internal load distribution of the power supply unit as close as possible and the boundary between the power supply units as clear as possible. Finally,based on the three-dimensional information of the actual load space distribution of a power grid,Kmeans and DPC clustering algorithms are used to verify that the proposed method effectively considers the actual spatial distribution of the loads. The internal load distribution of the divided power supply unit is close,and the boundaries between the divided power supply units are clear.

     

/

返回文章
返回