Paper Type

Complete

Paper Number

PACIS2025-1553

Description

The exponential rise in global energy demand which is driven by rapid urbanization and population growth has strained conventional power grids, necessitating smart and adaptive solutions. This paper presents a hybrid approach to short term load forecasting with smart grids, integrating machine learning (ML) and deep learning (DL) models to enhance management of energy and accuracy of prediction. This study evaluates artificial neural networks (ANN), Long Short-term memory (LSTM), and Time Augmented Transformer (TAT) models using real time data from UK Power Networks Open Data Portal. Additionally, a weighted bagging ensemble model (WBE) combines best performing models to achieve better prediction accuracy. Experimental results state that LSTM achieves highest predictive performance while the ensemble model offers a balanced trade-off between accuracy and robustness.

Comments

AI ML

Download

Share

COinS
 
Jul 6th, 12:00 AM

Optimizing Energy Management in Smart Grids: A Hybrid Approach to Load Forecasting

The exponential rise in global energy demand which is driven by rapid urbanization and population growth has strained conventional power grids, necessitating smart and adaptive solutions. This paper presents a hybrid approach to short term load forecasting with smart grids, integrating machine learning (ML) and deep learning (DL) models to enhance management of energy and accuracy of prediction. This study evaluates artificial neural networks (ANN), Long Short-term memory (LSTM), and Time Augmented Transformer (TAT) models using real time data from UK Power Networks Open Data Portal. Additionally, a weighted bagging ensemble model (WBE) combines best performing models to achieve better prediction accuracy. Experimental results state that LSTM achieves highest predictive performance while the ensemble model offers a balanced trade-off between accuracy and robustness.