The creation of a methodology for intelligent assessing and managing the security state of complex systems

Complex technical systems are the object of the study. The problem that is solved in the study is an increase in the level of security of complex technical systems. The originality of the study consists in: – comprehensive assessment of the security state of complex technical systems due to multi-level assessment using the theory of artificial intelligence; – reduced error in assessing the security state of a complex technical system due to the human factor due to the verification of the parameters of a complex technical system; – selection of the best individuals in bio-inspired algorithms, due to the use of an improved genetic algorithm, which achieves an increase in the efficiency and reliability of the obtained decisions and evaluations; – make accurate decisions by individually adjusting the actions of agents in each bio-inspired algorithm; – eliminating the conflict between agents in improved bio-inspired algorithms, which increases the efficiency and reliability of decisions made regarding the security state of complex technical systems; – implementation of deep learning of knowledge bases of agents of each bio-inspired algorithm, due to the method of deep learning, which achieves an increase in the efficiency and reliability of assessments and control effects on the security state of complex technical systems. Modeling of the proposed methodology was carried out, during which it was established that increasing the security of complex technical systems is achieved by increasing the efficiency of decision-making at the level of 15−17% due to the use of additional procedures and ensuring the reliability of decisions made at the level of 0.91. This study can be used in practice when taking into account the delay time for collecting and proving information from sensors (sensors) of complex technical systems.