基于Spark框架的电力调度异常数据辨识方法

Abnormal Data Identification Method of Power Dispatching Based on Spark Framework

由于电网运行数据具有多源、异构、高维等特征,使得传统检测方法已无法实现异常数据高效辨识,因此文章提出一种基于Spark框架的电力调度异常数据辨识方法.首先对电力调度数据进行聚类,然后在Spark框架下通过K-means算法对异常数据进行辨识,最后通过试验验证该方法的辨识效果.

Because the operation data of power grid has multiple sources,heterogeneous data and high dimension,the traditional detection method is unable to realize the efficient identification of abnormal data.Therefore,an identification method of abnormal data of power dispatching based on Spark framework is proposed.First,the power dispatch data were clustered,and then,the abnormal data were identified by the K-means algorithm under the Spark framework.Finally,the identification effect of the proposed method was verified by experiments.