基于word2vec的配电网恶意控制指令检测算法  

Malicious Control Command Detection Algorithm in Power Distribution Network Based on word2vec

作　　者：郑佩祥 陈彬 卢昕 徐文渊[2] ZHENG Peixiang;CHEN Bin;LU Xin;XU Wenyuan(State Grid Fujian Electric Power Company,Fuzhou 350003,China;College of Electrical Engineering,Zhejiang University,Hangzhou 310027,China)

机构地区：[1]国网福建省电力有限公司,福州350003 [2]浙江大学电气工程学院,杭州310027

出　　处：《计算机工程》2019年第4期119-123,129,共6页Computer Engineering

基　　金：国家高技术研究发展计划(2015AA050202);国家电网公司科技项目(52130415000P)

摘　　要：现有的配电网恶意控制指令检测方法基于电力系统运行规则,但规则维护困难、规则匹配耗时较长。根据配电网上行测量信息和下行控制指令之间存在的上下文一致性关系,提出基于word2vec的恶意控制指令检测算法。在配电网仿真平台模拟各类工况并获取标注数据集,结果表明,该算法能够取得100%的精确度和87.2%的召回率,具有较高的检测精度。The existing detection method of malicious control command of distribution network is based on the operating rules of power system,but the rules are difficult to maintain and the rule matching takes a long time.According to the context consistency relationship between the uplink measurement information of the distribution network and the downlink control instructions,a malicious control detection algorithm based on word2vec is proposed.Simulating various working conditions on the power distribution network simulation platform and obtaining the labeled data set,the results show that the algorithm can achieve 100 % accuracy and 87.2 % recall rate,and has high detection accuracy.

关 键 词：配电网 恶意控制指令 上下文 word2vec模型 异常检测 

分 类 号：TM769[电气工程—电力系统及自动化]