集成电力大数据日志分析模块的信息系统研究  

Research on Electric-Power Big Data Log Analysis Module in Information System

作　　者：余铮 冯浩 查志勇 YU Zheng;FENG Hao;ZHA Zhiyong(Information Communication Company,State Grid Hubei Electric Power Co.,Ltd.,Wuhan 430077)

机构地区：[1]国网湖北省电力有限公司信息通信公司,武汉430077

出　　处：《计算机与数字工程》2020年第3期571-574,733,共5页Computer & Digital Engineering

摘　　要：为了保障国家电网的信息系统高效运转,以及实现在电力大数据状态下的日志分析要求,该文提出一种基于K-最近邻算法的未知攻击识别算法。该算法可以高效地判断信息系统服务是否存在异常,关键文件是否被非法修改以及异常外联等行为。在该算法的基础上,实现了信息系统运行态势感知和全息智能处置系统。实验表明,该算法很好地支撑了系统的工作,同时也提高了日志分析的高效性。In order to ensure the efficient operation of information system,the state grid as well as the realization in the condition of electric power big data log analysis requirement,this paper puts forward a kind of unknown attack identification algorithm,based on K-nearest neighbor algorithm. This algorithm can effectively determine whether there is any abnormality in the information system service,whether the key files are illegally modified,and whether there is abnormal outreach. On the basis of this algorithm,the information system operation situation perception and holographic intelligent disposal system are realized. The experiment shows that the algorithm supports the work of the system and improves the efficiency of log analysis.

关 键 词：电力大数据 日志分析 异常检测 K-最近邻算法 未知攻击识别算法 

分 类 号：TP393.8[自动化与计算机技术—计算机应用技术]