Original scientific paper
https://doi.org/10.13044/j.sdewes.d13.0606
Machine Learning Electrical Load Forecasting: an application in microgrid energy consumption with adaboost regressor approach and a comparative study with hybrid method based on LSTM and MLP approaches
Yao Bokovi
; CERME/University of Lome, Lome, Togo
Kabe Moyème
; Ecole Nationale Supérieure d’Ingénieurs (ENSI), University of Lome, LOME, Togo
Sedzro Kwami Séname
; National Renewable Energy Laboratory, Golden, United States
Takouda Pidéname
; CERME/EPL, Lome, Togo
Lare Yendoubé
; CERME/University of Lome, Lome, Togo
Abstract
The dynamic evolution and variation of electrical loads is now, a priority for their optimal management and, above all, forecasting. Now, these dynamic load variations require computer tools that are able to implement optimal load forecasting models. Scientific research into automated models for forecasting electrical loads is therefore a challenge for scientific researchers, and several studies have been carried out in this area. These include machine learning approaches such as Long Short-Term Memory, Support Vector Machine, Multilayer Perceptron; deep learning, probabilistic and others. These studies are often quite complex due to the number of elevated hyperparameters they contain, with considerable deviations in accuracy between the real and predicted data. Thus, in order to exploit methods with fewer hyperparameters and minimized prediction deviations between consumed and production, this paper proposes a method for forecasting based on a regression ensemble method: adaboost regressor approach, to improve in energy consumption forecasting by application of advanced algorithm. So, this article presents learning and validation tests for the proposed model. The data used, were collected from a renewable energy source: photovoltaic solar energy. While 80% of the data collected was used for learning purposes, the remaining 20% was used for validation testing. The results of this study give a coefficient of determination R2 between 0.9995 and 0.9997 for the learning results and between 0.83 and 0.958 for the validation test results. According to the metrics parameters, these results are representative of the real data and reflect the performance of the proposed model. The proposed model is well adapted to the management of electrical consumption load forecasts to ensure balance between supply and demand.
Keywords
Optimal management; electricity demand; forecasting model; ensemble regression: adaboost regressor; LSTM; MLP
Hrčak ID:
342244
URI
Publication date:
18.5.2026.
Visits: 299 *