基于SSA-PFCM聚类算法的电力大数据异常检测

SSA-PFCM Clustering Algorithm-based Anomaly Detection from Electric Big Data

为提高电网数据异常检测可靠性,针对已有的PFCM算法存在的聚类中心难以选择的问题,提出了一种将PFCM算法和SSA相结合的电力大数据异常检测方法。其中,SSA算法被用来优化PFCM算法的初始中心。仿真验证表明改进后算法与传统算法相比具有优越性。

Aiming at improving the reliability of electric networks data detection and addressing the high difficulty of determining clustering center of the conventional possibilistic fuzzy c-means clustering method(PFCM),this work made a preliminary attempt to establish an anomaly detection method by combining PFCM and the sparrow search algorithm(SSA).The use of SSA was to optimize the initial center of PFCM.The modified algorithm was verified through comparative simulation to be superior to the conventional algorithm.