Observer-Based Resilient Adaptive Event-Triggered Control for Islanded Microgrids Under DoSAs and Unknown FDI Attacks

This study investigates the resilient event-triggered load frequency control (LFC) issue of islanded microgrids (MGs) subject to denial of service attacks (DoSAs) and unknown false data injection attacks (FDIAs) simultaneously. An augmented extended observer is established to estimate both unavailable system states and unknown FDIA signals. A resilient adaptive event-triggered (AET) scheme, incorporating DoSA detection, FDIA estimation, and dynamic threshold adjustment, is proposed to reduce network resource consumption while countering cyberattacks. Some sufficient criteria are derived to ensure the asymptotic stability of the augmented estimation error system with $H_{\infty } $ performance. Additionally, the uniformly ultimate boundedness of the system is also guaranteed even under hybrid cyberattacks. Finally, simulation studies are used to illustrate the effectiveness of the proposed method. Note to Practitioners—This study addresses the network-based load frequency control problem in isolated microgrids. Since microgrids typically rely on open communication networks to transmit control commands, they are vulnerable to cyberattacks, such as injected malicious signals or overwhelming request data. These attacks can result in significant frequency fluctuations. To tackle these challenges, this paper proposes an augmented extended observer capable of estimating both unknown injected signals and compromised system states. Furthermore, a resilient event-triggered control strategy is developed, utilizing the TCP protocol and estimated signals to counteract the effects of cyberattacks. This approach not only stabilizes frequency deviations under hybrid attack scenarios but also effectively reduces communication load. Notably, while this study focuses on load frequency control in isolated microgrids, the proposed resilient control algorithm is adaptable to other types of microgrids.