基于空间矢量变换与谱聚类的电压暂降事件数据空间聚合方法

Spatial aggregation method for voltage sags event data based on space vector transform and spectral clustering

随着用户对电能质量的要求越来越高,电网公司在变电站安装了大量电能质量监测装置。电网安装的监测装置越多,监测到的电压暂降事件也会越多,而采用监测到的电压暂降事件数量评价不同区域电网电压暂降水平,会导致评价结果失真,不利于电网公司准确认识各区域电网电能质量水平。为了消除不同区域电网安装的监测装置数量等因素的影响,准确评价不同区域电网真实电压暂降水平,对电压暂降事件进行聚合具有重要意义。为此,提出一种基于空间矢量变换与谱聚类的电压暂降事件数据空间聚合方法。首先,基于空间矢量变换方法提取持续时间与波形系数特征,消除了当变压器绕组联结方式不同时对电压暂降传播的影响。然后,提出自适应聚类簇数估计的谱聚类方法,以实现高效率的电压暂降数据空间聚合。最后,在IEEE 30节点系统中生成仿真数据进行验证,结果显示对于30条暂降数据能够在1.417 s内实现100%的聚合准确率,证明了该方法的正确性和高效性,具备实时在线应用的能力。

With the increasing demand of users for power quality,power grid companies have installed a large number of power quality monitoring devices in substations.The more monitoring devices are installed,the more voltage sag events are monitored,and the use of the number of monitored voltage sag events to evaluate the voltage sag level of different regional grids will lead to distorted evaluation results,which is not conducive to the grid company's accurate understanding of the power quality level of each regional grid.In order to eliminate the influence of factors such as the number of monitoring devices installed in different regional grids,and accurately evaluate the real voltage sag level of different regional grids,it is of great significance to aggregate voltage sag events.To this end,this paper proposes a spatial aggregation method of voltage sag event data based on space vector transform and spectral clustering.Firstly,based on the space vector transform method,the duration and waveform coefficient features are extracted to eliminate the influence of transformers with different winding connections on the propagation of voltage sag,and then a spectral clustering method with adaptive cluster number estimation is proposed to realize the spatial aggregation of voltage sag data with high efficiency.Simulation data are generated in the IEEE 30 nodes system for verification,and the results show that the aggregation accuracy of 100%can be realized in 1.417 seconds for 30 sag data,which proves the correctness and high efficiency of the method,and has the ability of real-time online application.