Analysis of statistical indicators of variance, asymmetry and excess in determining information security violations of cyberphysical systems of wind turbines

Abstract

The active introduction of intelligent systems that closely interact with physical processes to solve a wide range of different tasks of human life increases the relevance of risk analysis associated with the functioning of such systems. Such hybrid complex intelligent systems belong to the class of cyberphysical systems (CPS). Violations of CPS security caused by outside interference in the information flow can lead to economic losses, environmental threats, and threats to human life and health. A significant increase in incidents of violation of the safety of CPS wind turbines determines the relevance of research on methods for protecting such systems. The subject matter of the study in the article is the process of determining violations of the information security of the CPS of a wind generator based on the analysis of statistical indicators of variance, asymmetry, and kurtosis of the input parameter "Power" collected by CPS sensors. The goal is to develop an algorithm for determining violations of the information security of the CPS using methods for analyzing statistical indicators of variance, asymmetry, and kurtosis. The tasks to be solved are: to formalize the process of identifying falsified data in the information flow of the CPS; to determine the advantages and disadvantages of existing methods for ensuring the information security of the CPS; to determine the degree of changes in statistical indicators of variance, asymmetry, and kurtosis of the sample of the "Power" parameter of the wind generator in the presence of misinformation in the information flow; to analyze the possibility of supplementing and further improving the proposed algorithm. The methods used are analysis of statistical indicators of variance, asymmetry, and kurtosis of the sample of the parameter "Power" of the wind generator. The following results are obtained: the general characteristics of the CPS and features of the functioning of the CPS of the wind turbine as the object of research of this work are considered; an initial algorithm for determining violations of the information security of the CPS of wind turbine based on the use of statistical indicators of variance, asymmetry, and excess is developed; the fact of artificial substitution of data for the parameter "power" of the information flow of the CPS of a wind turbine is determined; ways to improve the developed algorithm using one-factor variance analysis, bootstrap methods are proposed. Conclusions. The scientific novelty of the results obtained consists of the development of an improved algorithm for determining the fact of data falsification in the CPS information flow based on the analysis of variance, asymmetry, and kurtosis indicators; the use of a statistical method for determining CPS security violations, analyzing the shortcomings of existing methods for determining CPS security violations and the possibility of their comprehensive improvement. The issues of the possibility of improving the developed method and testing the method in the future are also considered.

Keywords

• інформаційна безпека
• кфс
• статистичні показники
• дисперсія
• асиметрія
• ексцес