Iterative Learning Control for Load Frequency in Cyber-Attacked Multi-Area Power Systems

Sustaining the performance of the power grid at a desired operating point is a challenge in an uncertain environment. As a result of the environment’s dynamic and unpredictable nature, the multi-area-linked power system is impacted by communication limitations and load variance. Due to the widespread nature of electricity, several power regions share information inside a communication network. In complex networks, connectivity restrictions, which include time delays, information loss, and restricted bandwidth, may impair the dependability and scalability of the control method. Variations in the load can affect the stability of the power systems. Due to the active load change and the dynamic environment and uncertainty, a model-free controller is a suitable technique that can maintain reliable performance in this condition. The online adaptive policy control scheme is proposed for load frequency control (LFC) challenges in single and multi-area power grids. The proposed scheme uses an optimal control approach through a modified Bellman equation and two neural networks, one to approximate the proposed solving value function and the second one to approximate the optimal control. The performance of the proposed model free adaptive control is compared with that of the model predictive control