利用自组织映射SOM实现电力系统暂态稳定评估结果可视化

SOM based visualization of power system transient stability assessment results

自组织映射SOM(self-organizingmap)具有拓扑关系保持的优良性能,可应用于降维可视化。该文基于SOM实现暂态稳定评估结果的可视化。采用经典的SOM算法分别对原始特征和特征选取后的IEEE16机暂态稳定分类数据进行可视化,表明了可视化的可行性以及特征选取的重要性。采用一种改进的SOM算法—DPSOM算法,提高了可视化的性能,得到了满意的暂态稳定评估分类可视化结果。

Due to the topology-preserving nature,the SOM（self-organizing map） algorithm can be used to visualize the high-dimensional data.In this paper,the visualization of power system transient stability assessment results which based on SOM is realized.Firstly,for the IEEE 16 generators transient stability classified data, the original feature data and the feature selected data are visualized respectively by using the classic SOM algorithm,then the practicability of visualization and the importance of feature selection are explained. Finally, an improved SOM algorithm--DPSOM algorithm is adopted,which improves the performance of visualization and gets the satisfied results of visualization.