基于综合赋权和改进K-means++的DERs区域类型划分研究

Research on Regional Type Division of DERs Based on Comprehensive Weight and Improved K-means++

虚拟电厂和负荷聚合商能把分布式能源资源整合为资源聚合体,具有良好的发展前景。针对广泛分布式能源资源的城市区域,根据指标评价和聚类分析提出了一种区域类型划分方法。根据资源聚合体的特点,构建了能反映资源特性的综合指标体系;采用G1法和熵值法结合的综合赋权法确定各个指标权重;引入双尺度距离的概念,采用改进的K-means++算法对网格进行聚类分析,进而确定区域类型划分。最后,以北方某城市为例进行算例分析,验证了方法的适用性。

Virtual power plant and load aggregator has a good development prospect by integrating distributed energy resources(DERs)into resource aggregates.Aiming at the urban areas with widely distributed DERs,this paper proposes a method of regional type division based on index evaluation and cluster analysis.According to the characteristics of the resource aggregates,a comprehensive index system reflecting resource characteristics is constructed.The comprehensive weighting method combining with G1 method and entropy method is used to determine the weights of each index.The concept of dual-scale distance is introduced,and improved K-means++algorithm is used to cluster the grid,determining the region type division.Finally,an example of a northern city is given to verify the applicability of the method.