基于MMBN分群算法和容量加权参数聚合方法的风电场等值模型研究

Study on Wind Farm Equivalence Model Based on MMBN Clustering Algorithm and Capacity Weighting Parameter Aggregation Method

为提高风速分布不均匀时风电场等值模型的精度,提出了一种适用于双馈式风电场和鼠笼式风电场动态等值的方法。提出基于谱聚类的改进最大最小距离算法对风电场进行分群,该算法可以自动确定风电场最恰当的分群数量,避免传统算法分群数量的设置对仿真时间和精度的影响,提出一种可以无需改变变流器控制器参数的参数聚合方法,避免繁琐PI调整过程。通过PSCAD算例仿真及结果对比,表明该方法所建立的双馈式风电场和鼠笼式风电场等值模型的输出在稳态和故障情况下都能很好地表示详细模型。

In order to improve the accuracy of wind farm equivalent models under wind speed unequal distribution, a method is proposed for the wind farm dynamic equivalence with DFIG-based wind farm and FSIG-based wind farm. The improved max-min distance means based on spectral clustering algorithm is proposed to cluster the wind farm, and the number of the most appropriate clustering of wind farm is determined automatically. A parameter polymerization method without changing the control parameters of converter controller is applied to avoid the tedious PI adjustment process. By establishing models on PSCAD platform, the results show that the proposed equivalent model can more accurately describe the output of the detailed model under both steady state and transient state.