DETECTION OF CYBER ATTACKS ON POWER DISTRIBUTION SYSTEM USING QSVM

As cyber threats evolve, Power Distribution Systems (PDS) face growing risks from sophisticated attacks like False Data Injection Attacks (FDIAs), which can disrupt system stability and reliability. This thesis presents a quantum-based approach using Quantum Support Vector Machines (QSVM) to detect and mitigate FDIAs in PDS. By leveraging quantum feature mapping, the QSVM model efficiently identifies subtle anomalies within high-dimensional data, enhancing the accuracy and speed of FDIA detection. The methodology includes the integration of an augmented Lagrangian function to further optimize detection performance. Validated using the IEEE-13 bus system, this QSVM framework showcases its potential as a robust, real-time detection tool for cybersecurity in smart grid infrastructures. The results underscore the promise of quantum computing in strengthening the resilience of critical energy systems.