Prediction of Transient States in High-Voltage Power Grids Using Advanced Metaverse Digital Systems

Abstract

Transient states in high-voltage power grids represent critical challenges for power system operation and stability, where sudden load changes or faults lead to phenomena such as frequency transients, voltage fluctuations, and instability risks. This paper presents an advanced framework for predicting transient states using an integrated Metaverse environment that combines digital twin physical modeling, real-time simulation, and big data analysis through machine learning. The proposed system relies on comprehensive virtual simulation of high-voltage networks within an interactive Metaverse environment, integrating real-time and historical data from Phasor Measurement Units (PMUs), smart sensors, and SCADA systems. Deep learning algorithms (LSTM, Convolutional Neural Networks) are applied for predictive modeling, with augmented reality (AR) techniques for real-time visualization of predictions. Simulation results on the IEEE 39-bus system demonstrated a 34% improvement in prediction accuracy compared to conventional methods, with reduction of early detection time for critical states to less than 100 milliseconds. The study presents a practical framework for Metaverse applications in smart grid management and adaptive preventive control.