基于神经网络的变电站异常网络流量识别  

Identification of abnormal network traffic in substation based on neural network

作　　者：章杜锡 谢宏 李力 ZHANG Du-xi;XIE Hong;LI Li(State Grid Ningbo Electric Power Supply Company,Ningbo 315000,Zhejiang Province,China)

机构地区：[1]国网浙江省电力有限公司宁波供电公司,浙江宁波315000

出　　处：《信息技术》2020年第5期140-144,共5页Information Technology

摘　　要：针对变电站网络异常流量检测过程中需要兼顾准确度和实时性的问题,提出基于元启发优化算法的模块化的神经网络算法。仿真实验证明,在存在网络延迟和抖动的情况下,基于灰狼优化算法模块和进化策略算法模块的神经网络异常检测算法,能够准确和高效地识别变电站网络中的异常数据流量。Aiming at the problem of accuracy and real-time in the abnormal traffic detection process of substation network,a modular neural network algorithm based on meta-heuristic optimization algorithm is proposed.Simulation experiments show that the neural network anomaly detection algorithm based on gray wolf optimization algorithm module and evolution strategy algorithm module can accurately and efficiently identify abnormal data traffic in substation network in the presence of network delay and jitter.

关 键 词：智能变电站 神经网络 异常检测 

分 类 号：TM769[电气工程—电力系统及自动化]