基于模糊规则的电网工控系统异常识别方法研究  

Research on Abnormal Identification Method of Power Grid Industrial Control System Based on Fuzzy Rules

作　　者：韩立芝 王洪涛 朱兆群 王鹏朝 柏丽 李云山 Han Lizhi;Wang Hongtao;Zhu Zhaoqun;Wang Pengchao;Bai Li;Li Yunshan(State Grid Xinjiang Institute Co.,Ltd.,830002,China;State Grid Xinjiang Economic Research Institute,Urumqi,830002,China)

机构地区：[1]国网新疆电力有限公司,新疆乌鲁木齐830002 [2]国网新疆电力有限公司经济技术研究院,新疆乌鲁木齐830002

出　　处：《现代科学仪器》2024年第6期303-308,共6页Modern Scientific Instruments

摘　　要：由于电网工控系统中存在大量的不确定性和模糊性信息,一般的识别方法很难准确识别其异常问题,基于此,研究一种基于模糊规则的电网工控系统异常识别方法。通过模糊规则表建立关键特征与异常类别标签之间模糊规则,计算模糊特征与模糊规则条件部分之间的匹配度,利用Net Flow技术采集电网工控系统流量数据,采用递归式信息相关特征选择算法进行关键特征选取,从而推断出异常问题的类别。实验结果表明,研究方法能够准确识别电网工控系统中的异常问题,得出的Matthews相关性系数更大、更接近1。Due to the large amount of uncertainty and ambiguity information in the power grid industrial control system,general identification methods are difficult to accurately identify its abnormal problems.Based on this,a fuzzy rule-based method for identifying anomalies in power grid industrial control systems is studied.Establish fuzzy rules between key features and abnormal category labels through a fuzzy rule table,calculate the matching degree between fuzzy features and fuzzy rule conditions,use Net Flow technology to collect flow data from power grid industrial control systems,and use a recursive information related feature selection algorithm to select key features,thereby inferring the category of abnormal problems.The experimental results show that the research method can accurately identify abnormal problems in the power grid industrial control system,and the Matthews correlation coefficient obtained is larger and closer to 1.

关 键 词：模糊规则 电网 工控系统 Net Flow技术 关键特征 异常识别方法 

分 类 号：TP63.99[自动化与计算机技术—控制理论与控制工程]