A Multi-Layer Security Scheme for Mitigating Smart Grid Vulnerability against Faults and Cyber-Attacks

To comply with electric power grid automation strategies, new cyber-security protocols and protection are required. What we now experience is a new type of protection against new disturbances namely cyber-attacks. In the same vein, the impact of disturbances arising from faults or cyber-attacks should be surveyed by network vulnerability criteria alone. It is clear that the diagnosis of vulnerable points protects the power grid against disturbances that would inhibit outages such as blackouts. So, the first step is determining the network vulnerable points, and then proposing a support method to deal with these outages. This research proposes a comprehensive approach to deal with outages by determining network vulnerable points due to physical faults and cyber-attacks. The first point, the network vulnerable points against network faults are covered by microgrids. As the second one, a new cyber-security protocol named multi-layer security is proposed in order to prevent targeted cyber-attacks. The first layer is a cyber-security-based blockchain method that plays a general role. The second layer is a cyber-security-based reinforcement-learning method, which supports the vulnerable points by monitoring data. On the other hand, the trend of solving problems becomes routine when no ambiguity arises in different sections of the smart grid, while it is far from a big network’s realities. Hence, the impact of uncertainty parameters on the proposed framework needs to be considered. Accordingly, the unscented transform method is modeled in this research. The simulation results illustrate that applying such a comprehensive approach can greatly pull down the probability of blackouts.