Resilient Frequency Control of Microgrids Based on Descriptor-PI Observer Considering Nonlinearities and Multiple Sensors and Actuator Cyber-Attacks

Load frequency control (LFC) is essential in microgrids to maintain frequency stability by balancing power generation and demand. Ensuring the resilience of LFC against false data injection (FDI) attacks is crucial for reliable and secure operations. Therefore, this paper proposes a resilient and practical microgrid LFC framework designed to counteract coordinated multi-sensor and actuator FDI attacks, as well as load/generation disturbances, system uncertainties, and nonlinearities such as generation dead-bands (GDBs) and generation rate constraints (GRCs). First, a dynamic proportional-integral (DPI) observer-based descriptor approach has been designed for fast estimation of microgrid states and FDIs. Secondly, an adaptive Proportional-Integral-Derivative with a Type-2 Fuzzy controller (PID-T2Fuzzy) is developed, utilizing observer estimates to reject FDIs and disturbances. It adaptively tunes its gains through fuzzy mappings and employs a rate limiter for practical and stable operation. A Popov-Lyapunov stability analysis is conducted to assess the robust and stable performance of the LFC. The effectiveness of the proposed control framework is validated through simulations under various FDI scenarios, including state-dependent and state-independent attacks. The results demonstrate that the DPI observer provides fast and accurate state estimation, while its integration with the PID-T2Fuzzy controller ensures reliable frequency control under disturbances and FDIs. Furthermore, the scalability and real-time performance are verified in a three-area interconnected microgrid using hardware-in-the-loop (HIL) experiments.