Intelligent Anomaly Mitigation in Cyber-Physical Inverter-based Systems

TL;DR

This paper presents an AI-driven anomaly mitigation method using neural networks to enhance the security and reliability of inverter-based cyber-physical systems like microgrids, validated through real-time simulations.

Contribution

It introduces a novel data-driven AI approach employing neural networks for anomaly mitigation in inverter-based systems, improving resilience against cyber anomalies.

Findings

Effective anomaly mitigation demonstrated in microgrid control

Improved system robustness over traditional methods

Validated through real-time simulation results

Abstract

The distributed cooperative controllers for inverter-based systems rely on communication networks that make them vulnerable to cyber anomalies. In addition, the distortion effects of such anomalies may also propagate throughout inverter-based cyber-physical systems due to the cooperative cyber layer. In this paper, an intelligent anomaly mitigation technique for such systems is presented utilizing data driven artificial intelligence tools that employ artificial neural networks. The proposed technique is implemented in secondary voltage control of distributed cooperative control-based microgrid, and results are validated by comparison with existing distributed secondary control and real-time simulations on real-time simulator OPAL-RT.