Post-Quantum Secure Physics-Informed Federated Learning for Adaptive Protection in Inverter-Dominated Grids
Students & Supervisors
Student Authors
Supervisors
Abstract
"The increasing integration of inverter-based resources (IBRs) and renewable energy is reshaping modern distribution-grid operation, creating new challenges in protection coordination, frequency stability, and cybersecurity. Unlike conventional synchronous generators, inverter-interfaced sources contribute limited fault current and enable bidirectional power flow, reducing the effectiveness of classical relay-based protection schemes. In addition, fast-ramping loads from AI-driven infrastructure and large-scale data centres can create abrupt power imbalances in low-inertia systems. To address these challenges, this paper presents a post-quantum-ready physics-informed federated learning framework for adaptive protection and frequency support in inverter-dominated distribution grids. The proposed framework integrates physics-informed neural networks, federated learning, battery energy storage-based fast-frequency support, and a conceptual post-quantum secure communication layer. Simulation results on a modified IEEE 33-bus distribution system show that the proposed method improves disturbance classification accuracy, reduces false relay trips by 35–40%, and enhances frequency recovery under fast load-ramping conditions."
Keywords
Publication Details
- Type of Publication:
- Conference Name: "International Conference on Power, Electronics, Communications, Computing, and Intelligent Infrastructure 2026 "
- Date of Conference: 17/06/2026 - 17/06/2026
- Venue: Pabna University of Science and Technology (PUST), Pabna, Bangladesh
- Organizer: Pabna University of Science and Technology (PUST)