AI-driven forecasting for battery energy management in digital twin-integrated microgrids using machine learning techniques

دیجیتال توئین و ماشین لرنینگ

نوشته شده توسط سرکار خانم دکتر هما رشیدی زاده

منتشر شده در 21 مهر 1405

زمان مطالعه 60 دقیقه

دسته مقالات

The variability of renewable energy sources (RES) presents a challenge concerning the stability and operational efficiency of microgrids. This study suggests a Digital Twin-based framework encompassing deep learning and reinforcement learning (RL) techniques to improve the overall energy forecasting potential and optimization of battery management. Among the machine learning models tested, Deep Neural Network (DNN) turned out to be the most accurate and computationally efficient. When combined with RL, this allows the charging-discharging operation of batteries to be dynamically managed, thus maximizing energy efficiency and battery life. Through the implementation of this framework, microgrids will receive more reliable support via sustainable energy utilization and scalable support for intelligent energy management.

Please go to this link:

https://www.sciencedirect.com/science/article/abs/pii/S2352152X25042707