摘要
针对工控系统信息安全风险日益增加的问题,给出基于CNN-Focal入侵检测模型和状态监测防火墙的可行性解决问题方法,实现对工控系统的服务安全正常运行。试验表明:状态监测防火墙流量数据包丢包率低至16.6%,受攻击状态下工控系统吞吐量与正常情况下几乎相同;与支持向量机方法相比,基于CNN-Focal入侵检测模型的损失函数综合指标更好。
Aiming at the problem of increasing information security risks of industrial control systems,a feasible solution method based on the CNN-Focal intrusion detection model and state monitoring firewall was given to realize the safe and normal operation of the services of the industrial control system.The experiments show that the packet loss rate of traffic data in a condition monitoring firewall is as low as 16.6%,and the throughput of the industrial control system under attack is almost the same as that under normal conditions;compared with the support vector machine method,the comprehensive index of loss function based on the CNN-Focal intrusion detection model is better.
作者
吴金宇
陶文伟
吴昊
陈刚
Wu Jinyu;Tao Wenwei;Wu Hao;Chen Gang(China Southern Power Grid Co., Ltd., Guangzhou Guangdong 510623,China;China Southern Power Grid Digital Grid Research Institute Co., Ltd. ,Guangzhou Guangdong 510555, China)
出处
《电气自动化》
2021年第3期92-95,共4页
Electrical Automation
关键词
工业控制系统
电网工控系统
安全运行
可信网络
深度学习
industrial control system
power grid industrial control system
safe operation
trusted network
deep learning