SCADA Digital Twin and Cryptographic AMI Cyber-Twin for Smart Grid Security

Smart grids increasingly depend on tightly coupled Supervisory Control and Data Acquisition (SCADA) and Advanced Metering Infrastructure (AMI) infrastructures to enable real-time monitoring, control, and data-driven energy management. Yet, the growing connectivity of 5G/6G-enabled systems expands the cyberattack surface and intensifies risks of SCADA component disruption and AMI data leakage. This study proposes a dual-twin framework (SCADA Digital-Twin with AMI Cyber-Twin) that improves both SCADA vulnerability intelligence and secure AMI-based energy forecasting. We propose a new algorithm that adopts the Directed Graph (DiGraph) Digital Twin mechanism with Label-encoded Shallow Machine Learning (DDT-LE-SML) to localize cyberattack-induced vulnerabilities and support attack classification in the AMI of the SCADA system. The proposed DDT-LE-SML algorithm works with Long-Short-Term-Memory (LSTM)-based FedAggSum-PredDF and Cyber-Twin Fernet-AES-HMAC strategy. Each physical smart meter is mirrored by a virtual Cyber-Twin node to enable encrypted handling of prediction data and privacy-preserving aggregation of distributed energy load forecasts. Experimental evaluation demonstrates higher overall accuracy ≈ (98%) and stable, encrypted distributed forecasting across the Cyber-Twin node, demonstrating stable encrypted distributed forecasting across Cyber-Twin nodes while enabling effective SCADA vulnerability localization. The proposed framework therefore supports risk-aware monitoring and privacy-preserving energy intelligence in next-generation smart-grid environments.