改进k-means的电网控制自动化系统数据聚类方法

Improved k-means Data Clustering Method for Power Grid Control Automation System

为了解决海量数据导致电网控制效率低的问题,提出改进k-means的电网控制自动化系统数据聚类方法。首先,利用主成分分析法提取出电网控制自动化系统数据主成分,将数据简化和降维;然后,通过滤波去噪处理电网数据,保证聚类样本干净整洁;最后,在改进k-means聚类算法下对数据聚类,确定k值得到最优聚类结果,实现电网控制自动化系统数据聚类。实验结果,对比每组样本下3种方法的Silhouette值,每组实验中Silhouette值的最大值都是所提方法,说明所提方法的聚类难度小和频率控制稳定性小。

In order to solve the problem of low power grid control efficiency caused by massive data,a data clustering method of power grid control automation system based on improved k-means is proposed.Firstly,the principal component analysis method is used to extract the data principal component of power grid control automation system,simplify and reduce the dimension of the data.Then,the power grid data is processed by filtering and denoising to ensure that the clustering samples are clean and tidy.Finally,the data is clustered under the improved k-means clustering algorithm,the k value is determined,the optimal clustering result is obtained,and the data clustering of power grid control automation system is realized.The experimental results compare the Silhouette values of the three methods under each group of samples.The maximum value of Silhouette value in each group of experiments is the method proposed,which shows that the proposed method is less difficult to cluster and strong stable in frequency control.