基于支持向量法的风电场概率聚类等值方法  

Probabilistic Aggregation of Wind Farm Based on Support Vector Clustering Method

作　　者：常方圆 郑良 丁伟[1] CHANG Fang-yuan;ZHENG Liang;DING Wei(International Center of Small Hydropower,Hangzhou 310002,China)

机构地区：[1]国际小水电中心,浙江杭州310002

出　　处：《浙江水利水电学院学报》2022年第1期82-85,共4页Journal of Zhejiang University of Water Resources and Electric Power

基　　金：浙江省科技计划项目(2019C04019)。

摘　　要：介绍并验证一种基于支持向量法的概率聚类风电场等值建模方法,该方法能够在一年的系统研究中得出用于表征大型风电场的等值风机数量。根据风力分布情况以及流入风速的大小对风电机组进行聚类并分组,通过分析某点风力情况来确定表征风电场的等值风机的数量。通过一个包含9台风电机的13.5MW风电场接入电网验证此方法,在给定风力条件下,聚合风电场模型的动态响应与风电场全模型的动态响应进行对比,得出最终结论。此方法亦可作为构建风电并网物理/数字混合仿真实验平台的建模基础。A probabilistic aggregation of wind farm based on support vector clustering method is introduced,which can obtain the number of equivalent wind turbines used to characterize large-scale wind farms in one year of system research.The wind turbines are clustered and grouped according to the wind distribution and the inflow wind speed,and the number of equivalent wind turbines characterizing the wind farm is determined by analyzing the wind situation at a certain point.This method is verified by a13.5 MW wind farm containing 9 wind turbines,which is connected to the grid.Under given wind conditions,the dynamic response of the wind farm model is aggregated with the dynamic response of the full wind farm model,and the final conclusion is obtained,which can also be used as the modeling basis for building a wind power physics/digital hybrid simulation experimental platform.

关 键 词：风力发电 风电场建模 概率聚类 支持向量法 

分 类 号：TH39[机械工程—机械制造及自动化]