An Industrial Internet Security Assessment Model Based on a Selectable Confidence Rule Base

Abstract

To mitigate the impact of network security on the production environment in the industrial internet, this paper proposes a confidence rule-based security assessment model for the industrial internet that uses selective modeling. First, a definition of selective modeling tailored to the characteristics of the industrial internet is provided. Based on this, the assessment process of the Selectable Belief Rule Base (BRB-s) model is introduced. Then, in combination with the Selection covariance matrix adaptive evolution strategy (S-CMA-ES) algorithm, a parameter optimization method for the BRB-s model is designed, which expands the selective constraints on expert knowledge. This model establishes a better unidirectional selection strategy among different subgroups, and while expanding the selection constraints on expert knowledge, it achieves better evaluation results. This effectively addresses the issue of reduced modeling accuracy caused by insufficient data and poor data quality. Finally, the experiments of different evaluation models on industrial data sets are compared, and good results are obtained, which verify the evaluation accuracy of the industrial Internet network security situation assessment model proposed in this paper and the feasibility and effectiveness of the S-CMA-ES optimization algorithm.