Distributed Micro-Grid Attack Elasticity for Frequency Stability

The patent implements attack prediction by constructing a dynamic equation model of the micro-grid system and using it to predict future system states. This preliminary action allows the control system to anticipate potential attack effects before they fully manifest, enabling proactive defense measures to be taken. The prediction mechanism calculates expected state sequences and compares them with actual measurements to detect anomalies indicating attacks.

The patent introduces an attack elasticity control module as an intermediary between the traditional hierarchical control architecture and the external environment. This intermediary layer includes a prediction mechanism and auxiliary control input that mediates the system's response to potential attacks, filtering malicious influences before they can disrupt the core droop control functionality while maintaining normal operation during non-attack conditions.

The patent segments the control system into distinct functional modules: a prediction mechanism that constructs dynamic equations and predicts future states, an attack detection module that compares predicted states with actual measurements, and an auxiliary control module that generates correction inputs. This segmentation allows each module to perform specialized calculations efficiently, distributing the computational burden across modular components rather than concentrating it in a single complex controller.

The patent implements partial action by selectively applying attack correction only when anomalies are detected. The system continuously monitors using lightweight prediction but only activates the full auxiliary control correction mechanism when the comparison between predicted and actual states indicates an attack. This approach reduces average calculation burden while maintaining security when needed.