System for discovering abnormal behaviours of industrial control system based on neural network and optimal recommendation

Abstract

The invention belongs to the technical field of information, and discloses a system for discovering the abnormal behaviours of an industrial control system based on a neural network and optimal recommendation. Flow data of the industrial control system is analyzed; dimensionality reduction of high-dimensional data in the industrial control system is carried out; irrelevant attribute deletion and redundant attribute redundancy elimination are realized; dynamic selection of intrusion features is realized while the internal relation of real input and output samples is followed; the system performs modelling for detected flow data in different scenes; and the sample flow and the flow after modelling is successful are compared, so that the difference of the detected flow data features and normal flow features is detected, so that abnormal flow is discovered. According to the system for discovering the abnormal behaviours of the industrial control system based on the neural network and optimal recommendation disclosed by the invention, scene comparison and mining analysis of flow are carried out based on the optimal recommendation theory; an optimal recommendation model of malicious flow (scene) is established; and in addition, by utilization of a cloud computing technology, malicious attack behaviour discovery and detection functions of customized industrial control equipment and systems based on a cloud platform are designed and realized.

Description

technical field [0001] The invention belongs to the field of information technology, and in particular relates to a system for discovering abnormal behaviors of industrial control systems based on neural networks and optimal recommendations. Background technique [0002] With the development of computer technology, communication technology and control technology, the traditional industrial control field is undergoing an unprecedented change, and it is beginning to develop in the direction of networking. The structure of the industrial control system has evolved from the initial CCS (Computer Centralized Control System), to the second generation of DCS (Distributed Control System), to the now popular FCS (Fieldbus Control System); linking networking with Fieldbus . The more influential field bus systems in the field of industrial control include: FF, LonWorks, Profibus, CAN, HART, and RS485 bus network. The unified standard (FF) that has been formulated by the Fieldbus Foun...

AI Technical Summary

Problems solved by technology

[0004] To sum up, the problems existing in the existing technology are: the traditional detection system has the characteristics of low detection rate and slow detection speed when dealing with irrelevant attributes and redundant attributes in high-dimensional data.

Images