基于深度自回归模型的电网异常流量检测算法  

Abnormal flow detection algorithm for power grid based on deep autoregressive model

作　　者：李勇[1] 韩俊飞[1] 李秀芬[1] 王鹏[1] 王蓓[1] LI Yong;HAN Junfei;LI Xiufen;WANG Peng;WANG Bei(Institute of Information and Communication Technology,Inner Mongolia Electric Power Research Institute,Hohhot 010020,Inner Mongolia,China)

机构地区：[1]内蒙古电力科学研究院信息通信技术研究所,内蒙古呼和浩特010020

出　　处：《沈阳工业大学学报》2024年第1期24-28,共5页Journal of Shenyang University of Technology

基　　金：国家自然科学基金项目(61033013)。

摘　　要：针对电网中行为种类复杂多样且数量众多的问题,提出了一种基于自回归模型的电网异常流量检测算法。该算法利用深度自编码网络自动提取网络流量数据的特征,降低异常流量检测的分析周期,并自动挖掘数据的层次关系。通过支持向量机对提取的特征进行分类,实现对异常流量的检测。仿真实验结果表明,所提算法可以分析不同攻击向量,避免噪声数据的干扰,进而提高电网异常流量检测的精度,对于流量数据处理具有重要意义。A power grid abnormal flow detection algorithm based on autoregressive model was proposed to address the complex and diverse behavior types and numerous issues in the power grid.Deep auto-encoding networks were used to automatically extract the features of network flow data,reduce the analysis cycle of abnormal flow detection,and automatically mine the hierarchical relationships of the data.At the same time,support vector machine was used to classify the extracted features and detect abnormal flow.Simulation experiment results show that this algorithm can analyze different attack vectors,avoid interference from noisy data and improve the accuracy of abnormal flow detection in the power grid.This method has high anti-interference ability and detection accuracy,and is of great significance for flow data processing in large-scale power grids.

关 键 词：自回归模型 深度学习 异常检测 海量数据 分析周期 支持向量机 

分 类 号：TM76[电气工程—电力系统及自动化]