Enhancing the Power Load Prediction using LSTM in a Smart Grid Scenario for Household Data

Le résumé fourni par la source

The smart grid fuses modern technology with traditional infrastructure to enhance efficiency and sustainability. Accurate Load Forecasting (LF) is crucial for ensuring reliability and rising electricity demand, while integration of machine learning transforms energy and grid management. In this research real-time consumption of historical power load data from house hold smart meters is used. The aim of this paper is to choose the best machine learning algorithm with good accuracy which helps to predict the future power load and assess the load forecasts with low Root Mean Squared Error (RMSE) and Mean Absolute Percentage Error (MAPE) and high R2Score. Our research involved the implementation of Machine Learning (ML) algorithms using load forecasting. This paper results the effectiveness and efficiency of ML algorithms in predicting future power loads, with LSTM demonstrating low RMSE, MAPE, of 1.40 and 0.03(%) and R2Score of 0.909 respectively. Overall, the integration of ML algorithms in load forecasting for smart grids shows great potential for enhancing efficiency and reliability in energy management.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.