Defending against false data injection attacks is critical to ensuring the stable operation of DC microgrids (MGs). Based on the traditional linear extended state observer (LESO), this article proposes a nonlinear extended state observer (NESO). While the conventional LESO is applicable to low-frequency attack scenarios, the proposed NESO, integrated with a predefined-time algorithm, enhances the dynamic performance near the steady-state. This enables it to effectively mitigate a wide range of cyber-attacks, including high-frequency ones. Experimental results demonstrate that under attacks involving constant, sinusoidal, and noise signal injections, the NESO can rapidly and accurately track and compensate for the attack signals, thereby safeguarding the stable operation of the DC MGs system.
Ding et al. (Fri,) studied this question.