Shovon, Md Shihab Sadik and Hasan, A S M Mahamudul and Subha, Dil Tabassum and Zubair, Sk Md (2026) Physics-Informed AI-Driven Digital Twin Framework for Renewable Energy Forecasting and Real-Time Power Grid Integration. American Journal of Technology Advancement, 3 (8). ISSN 2997-9382

Text AJTA_Physics-Informed+AI-Driven.pdf Download (675kB)

Abstract

The current study proposes an innovative physics-informed AI-driven digital twin methodology, which is expected to improve the forecasting of renewable energy generation and allow for the safe, flexible, and efficient integration of renewable sources in modern power grids. The suggested framework integrates the knowledge of artificial intelligence and physical principles that can be helpful in overcoming the challenges associated with the variable nature of renewable generation, especially that of solar and wind sources, whose operation is heavily dependent on the surrounding environment. Secondary research and exploratory data analysis performed using a large time-series dataset allowed for identifying patterns in the temporal generation of renewables, correlation between renewables and environmental parameters, dynamics of changes in the demand, behavior of batteries, and operational conditions of grids. It was found that there are substantial dependencies of renewable generation from environmental conditions, permanent imbalances between supply and demand, and an important function of battery systems in maintaining the resilience of grids. The review of literature confirmed the need for hybrid methodologies, which take into account both forecasting accuracy and physical consistency. Digital twin technology allows for representing the dynamic state of power grids.

Item Type:	Article
Uncontrolled Keywords:	Physics-Informed AI, Digital Twin, Renewable Energy Forecasting, Power Grid Integration, Smart Grid
Subjects:	H Social Sciences
Depositing User:	admin eprints
Date Deposited:	03 Sep 2026 11:58
Last Modified:	03 Sep 2026 11:58
URI:	http://eprints.umsida.ac.id/id/eprint/17044

Actions (login required)

View Item