Ensuring Cyber Security in Secondary Voltage Regulation in Multi-Agent Control Systems of Cyber-Physical Microgrids

The article addresses a study of the effectiveness of distributed multi-agent systems (MAS) for managing microgrids and electric networks with renewable energy sources with an emphasis on ensuring resistance to cyber threats. The purpose of the study is to develop an efficient approach to detecting and eliminating the consequences of cyberattacks through probabilistic analysis of agent behavior and algorithms for restoring data quality during secondary voltage regulation. The methods include machine learning techniques and probabilistic models to optimize control and stabilize network operation modes. The article considers various structures for control of cyber-physical microgrids based on MAS and analyzes possible cyberattacks on MAS management. The influence of cyberattacks on the MAS topology and on the consistency of agents' actions during secondary voltage regulation is shown. An approach based on probabilistic analysis and machine learning techniques has been developed that makes it possible to timely respond to cyber threats, eliminate their consequences and follow a strategy for optimal management of cyber-physical microgrids. A strategy is proposed that increases the efficiency of intelligent management and maintains the stability of energy flows even when there are information security risks.