ICPS入侵检测深度学习混合模型  

Deep Learning Hybrid Model for ICPS Intrusion Detection

作　　者：金浩 孙子文 JIN Hao;SUN Zi-wen(School of Internet of Things,Jiangnan University,Wuxi 214122,China;Engineering Research Center of Internet of Things Technology Applications Ministry of Education,Jiangnan University,Wuxi 214122,China)

机构地区：[1]江南大学物联网工程学院,江苏无锡214122 [2]江南大学物联网技术应用教育部工程研究中心,江苏无锡214122

出　　处：《控制工程》2021年第8期1708-1716,共9页Control Engineering of China

基　　金：国家自然科学基金资助项目(61373126);中央高校基本科研业务费专项资金资助项目(JUSRP51510)。

摘　　要：针对工业信息物理系统面临的网络安全问题,研究了一种基于深度学习混合模型的入侵检测方案。该方案将基于深度信念网络的无监督学习策略与基于支持向量机的有监督学习策略相结合,以实现工业信息物理系统入侵检测的半监督学习。对原始数据进行归一化处理并采用深度信念网络进行数据降维后,利用支持向量机进行入侵检测。使用MATLAB工具进行仿真,对以Modbus作为通信协议的监控与数据采集系统的真实数据进行测试。结果表明,与深度信念网络、支持向量机等算法模型相比,深度学习混合模型能显著提高异常检测的准确度。Aiming at the network security problem faced by industrial cyber physical systems(ICPS),an intrusion detection scheme based on the deep learning hybrid model is studied.The scheme combines the unsupervised learning strategy based on the deep belief network with the supervised learning strategy based on the support vector machine to realize semi-supervised learning of industrial cyber physical system intrusion detection.After the original data are normalized and the dimensionality of the data is reduced by the deep belief network,the support vector machine is used for intrusion detection.The MATLAB tool is used for simulation to test the real data of the supervisory control and data acquisition system with Modbus as the communication protocol.The simulation results show that compared with the models of the deep belief network,support vector machine and other algorithms,the deep learning hybrid model can significantly improve the accuracy of anomaly detection.

关 键 词：工业信息物理系统 深度信念网络 支持向量机 半监督 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]