Industrial Control System Modeling Based on Knowledge Graph

Jing Teng,Ruge Yang,Tian Wang,Hetian Feng,Yaofu Cao,Junlu Yan

DOI: https://doi.org/10.1109/cac59555.2023.10451881

2023-01-01

Abstract:In the context of “Industry 4.0”, the integration of industrialization and informatization has deepened. Industrial Control System (ICS) is more based on heterogeneous data and knowledge integration for cyber-physical production, and the openness and integration make ICS more vulnerable to attacks. Since ICS directly controls physical equipment, a breach could cause an irreversible disaster. It is especially important to have a security defense for ICS. Existing research has found that modeling the cyber and physical information of ICS helps security defense. Knowledge graphs, as an important technique of knowledge modeling, can mine the potential relationship between data and organize information effectively. Modeling ICS with knowledge graphs suits its data-intensive and relationship-rich requirements. This work proposes a knowledge graph-based information modeling for ICS, which focuses on ICS's monitoring and control layers and considers the relationships between network and physical devices. The model provides support for the subsequent safe operation and defense of ICS. A specific example study of the Secure Water Treatment for water treatment systems illustrates the potential for modeling ICS using knowledge graphs.