KPCA-IPSO-OCSVM方法在工业控制系统入侵检测中的应用
首发时间:2018-11-06
摘要:为了提高工业控制系统入侵检测的准确性,本文面向Modbus TCP协议的工控系统提出一种基于KPCA-IPSO-OCSVM算法的入侵检测方法。首先,采用核主成分分析(Kernel Principal Component Analysis, KPCA)对强非线性、高复杂度和高维度的工业数据进行特征提取,消除冗余特征、降低数据维度;其次,采用免疫粒子群(Immune Particle Swarm Optimization,IPSO)优化单类支持向量机(One Class Support Vector Machine, OCSVM)构建更准确的入侵检测模型。在实验室建立仿真环境,模拟工业控制系统运行场景,获取实际数据,验证设计方法的可行性和优越性。结果论证了该设计方法可以精确甄别异常行为,提升入侵检测准确性和工控系统安全性。
关键词: 工业控制系统 入侵检测 KPCA IPSO OCSVM Modbus TCP协议
For information in English, please click here
Application of KPCA-IPSO-OCSVM Algorithm in Intrusion Detection of Industrial Control System
Abstract:In order to improve the accuracy of industrial control system intrusion detection, this paper proposes an intrusion detection method based on KPCA-IPSO-OCSVM algorithm for the industrial control system of Modbus TCP protocol. Firstly, the kernel principal component analysis (KPCA) is used to extract features of strong nonlinear, high complexity and high dimensional industrial data, eliminating redundant features and reducing data dimensions. Secondly, using the immune particle swarm optimization (IPSO) to optimize one class support vector machine (OCSVM) algorithm can obtain a more accurate intrusion detection model. By establishing a simulation environment in the laboratory, simulating the operating scenarios of the industrial control system to obtain experimental data and verify the feasibility and superiority of the design method.
Keywords: Industrial Control System Intrusion Detection KPCA IPSO OCSVM Modbus TCP Protocol
基金:
引用
No.****
动态公开评议
共计0人参与
勘误表
KPCA-IPSO-OCSVM方法在工业控制系统入侵检测中的应用
评论
全部评论0/1000