Enhancing Fault and Cyberattack Detection in Power Systems Using Digital Twin and SVM

Detecting cyberattacks in power systems is critical to preventing disruptions, ensuring grid stability, and protecting against malicious manipulations that could lead to cascading failures, financial losses, and threats to national security. This paper presents a digital twin (DT) and support vector machine (SVM)-based framework for enhanced fault detection and cyberattack mitigation in power systems. The IEEE 33-Bus distribution system is modeled in MATLAB/Simulink to generate real-time datasets, enabling the SVM classifier to accurately distinguish normal operations, electrical faults, and false data injection (FDI) attacks. By leveraging real-time DT-generated data, the proposed framework improves classification accuracy and reduces false positives. Simulation results validate its effectiveness, demonstrating the potential of the DT-SVM framework for real-time smart grid security and adaptive fault detection.