Detection and mitigation of coordinated cyber-physical attack in CPPS.

Area of Science:

• Electrical Engineering
• Computer Science
• Cybersecurity

Background:

• Cyber-Physical Power Systems (CPPS) integrate internet elements with physical power grids, improving monitoring and control.
• CPPS are vulnerable to cyberattacks due to inadequate cybersecurity, necessitating adaptive defense strategies.
• Conventional offline controllers struggle with the dynamic operating conditions of modern CPPS.

Purpose of the Study:

• To develop and evaluate an adaptive defense strategy for analyzing and mitigating coordinated attacks in CPPS.
• To propose a novel framework integrating STATCOM-based Adaptive Model Predictive Controller (AMPC) with RPME and time delay compensation.
• To assess the impact, detection, and mitigation of physical, cyber, and coordinated attacks in CPPS.

Main Methods:

• A framework combining STATCOM-based AMPC with RPME and time delay compensation was developed.
• Data-driven anomaly detection methods including Convolutional Neural Network (CNN), Support Vector Machine (SVM), Random Forest (RF), and K Nearest Neighbour (KNN) were employed.
• Case studies were conducted using time and frequency domain simulations on the WSCC 3 machine 9 bus system.

Main Results:

• CNN demonstrated superior performance in detecting anomalies in PMU data compared to SVM, RF, and KNN.
• The proposed STATCOM-based AMPC framework with RPME achieved faster system recovery from coordinated attacks than conventional controllers.
• The strategy effectively mitigated coordinated attacks, validating its efficacy on a standard power system model.

Conclusions:

• The proposed adaptive defense strategy effectively addresses the vulnerabilities of CPPS to coordinated cyberattacks.
• CNN is a highly effective tool for anomaly detection in CPPS, crucial for identifying cyber threats.
• The integrated STATCOM-based AMPC framework offers a robust solution for enhancing the resilience and stability of power systems against sophisticated attacks.