Improved Resilient Model Predictive Control for Enhanced Microgrid Virtual Inertia Emulation by Virtual Energy Storage System Under DoS Attacks

Abstract

The distributed control of a microgrid is fully dependent on advanced information and communication technologies that are sensitive to cyber-physical systems. Cyberattacks, such as denial-of-service (DoS) attacks, can cause unstable operation of low-inertia microgrids. This paper proposes enhanced microgrid virtual inertia control under DoS attacks using an improved resilient model predictive control (IRMPC)-based virtual energy storage system (VESS). IRMPC comprises an attack detector, an autoregressive (AR)-based signal estimator, and an MPC-based VESS controller. An attack detector was used to detect the DoS attacks. An AR-based signal estimator is then used to estimate the feedback data that are subjected to DoS attacks. The firefly algorithm was used to optimize the AR parameters. The effectiveness of the proposed IRMPC was compared with that of conventional model predictive control, conventional model predictive control-based VESS, and resilient model predictive control-based VESS. The simulation results revealed that under a DoS attack, the proposed IRMPC can successfully improve the microgrid virtual inertia emulation. Additionally, the proposed IRMPC has a performance effect over the compared techniques in terms of the reduction in RoCoF deviation and frequency deviation during normal situations, DoS attacks, and disconnection of wind turbine generation. The simulation results also confirmed that IRMPC is robust to microgrid parameter variations when compared to the other methods.

Keywords

• Virtual inertia emulation
• resilient model predictive control
• DoS attacks
• microgrid
• virtual energy storage system